[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14063502#comment-14063502 ] Hudson commented on YARN-1408: -- FAILURE: Integrated in Hadoop-Hdfs-trunk #1806 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk/1806/]) YARN-1408 Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins. (Sunil G via mayank) (mayank: http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1610860) * /hadoop/common/trunk/hadoop-yarn-project/CHANGES.txt * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/RMContainer.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/RMContainerImpl.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/AbstractYarnScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/AppSchedulingInfo.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/SchedulerApplicationAttempt.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CapacityScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/common/fica/FiCaSchedulerApp.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FSSchedulerApp.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FairScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/TestRMContainerImpl.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestCapacityScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FairSchedulerTestBase.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/TestFairScheduler.java > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: YARN-1408-branch-2.5-1.patch, Yarn-1408.1.patch, > Yarn-1408.10.patch, Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.8.patch, Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.se
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14063473#comment-14063473 ] Hudson commented on YARN-1408: -- FAILURE: Integrated in Hadoop-Mapreduce-trunk #1833 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk/1833/]) YARN-1408 Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins. (Sunil G via mayank) (mayank: http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1610860) * /hadoop/common/trunk/hadoop-yarn-project/CHANGES.txt * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/RMContainer.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/RMContainerImpl.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/AbstractYarnScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/AppSchedulingInfo.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/SchedulerApplicationAttempt.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CapacityScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/common/fica/FiCaSchedulerApp.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FSSchedulerApp.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FairScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/TestRMContainerImpl.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestCapacityScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FairSchedulerTestBase.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/TestFairScheduler.java > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: YARN-1408-branch-2.5-1.patch, Yarn-1408.1.patch, > Yarn-1408.10.patch, Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.8.patch, Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hado
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14063361#comment-14063361 ] Hudson commented on YARN-1408: -- FAILURE: Integrated in Hadoop-Yarn-trunk #614 (See [https://builds.apache.org/job/Hadoop-Yarn-trunk/614/]) YARN-1408 Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins. (Sunil G via mayank) (mayank: http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1610860) * /hadoop/common/trunk/hadoop-yarn-project/CHANGES.txt * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/RMContainer.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/RMContainerImpl.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/AbstractYarnScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/AppSchedulingInfo.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/SchedulerApplicationAttempt.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CapacityScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/common/fica/FiCaSchedulerApp.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FSSchedulerApp.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FairScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/TestRMContainerImpl.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestCapacityScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FairSchedulerTestBase.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/TestFairScheduler.java > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: YARN-1408-branch-2.5-1.patch, Yarn-1408.1.patch, > Yarn-1408.10.patch, Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.8.patch, Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.serv
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14062879#comment-14062879 ] Mayank Bansal commented on YARN-1408: - Committed to trunk, branch 2 and branch-2.5. branch-2.5 needed some rebase , Updating the patch for branch-2.5 Thanks, Mayank > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: YARN-1408-branch-2.5-1.patch, Yarn-1408.1.patch, > Yarn-1408.10.patch, Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.8.patch, Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14062716#comment-14062716 ] Hudson commented on YARN-1408: -- FAILURE: Integrated in Hadoop-trunk-Commit #5887 (See [https://builds.apache.org/job/Hadoop-trunk-Commit/5887/]) YARN-1408 Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins. (Sunil G via mayank) (mayank: http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1610860) * /hadoop/common/trunk/hadoop-yarn-project/CHANGES.txt * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/RMContainer.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/RMContainerImpl.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/AbstractYarnScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/AppSchedulingInfo.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/SchedulerApplicationAttempt.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CapacityScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/common/fica/FiCaSchedulerApp.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FSSchedulerApp.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FairScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/rmcontainer/TestRMContainerImpl.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestCapacityScheduler.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/FairSchedulerTestBase.java * /hadoop/common/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/fair/TestFairScheduler.java > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch, > Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, > Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, > Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcon
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14062698#comment-14062698 ] Mayank Bansal commented on YARN-1408: - +1 Committing Thanks [~sunilg] for the patch. Thanks [~jianhe], [~vinodkv] and [~wangda] for the reviews. Thanks, Mayank > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch, > Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, > Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, > Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14062696#comment-14062696
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12655718/Yarn-1408.11.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:red}-1 core tests{color}. The patch failed these unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager:
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServices
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesCapacitySched
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesAppsModification
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesNodes
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesApps
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesFairScheduler
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4312//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4312//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch,
> Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch,
> Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch,
> Yarn-1408.9.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14061872#comment-14061872
]
Hadoop QA commented on YARN-1408:
-
{color:green}+1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12655718/Yarn-1408.11.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 core tests{color}. The patch passed unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager.
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4303//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4303//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch,
> Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch,
> Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch,
> Yarn-1408.9.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14061549#comment-14061549 ] Karthik Kambatla commented on YARN-1408: Quickly glanced through the patch. Changes seem reasonable. Given Jian, Vinod and others have already reviewed this closely, I haven't done a detailed review. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch, > Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, > Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, > Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14061540#comment-14061540 ] Karthik Kambatla commented on YARN-1408: Let me take a quick look at this as well. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch, > Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, > Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, > Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14061337#comment-14061337 ] Jian He commented on YARN-1408: --- one minor comment while investigating test failures, visibility for AbstractYarnScheduler#recoverResourceRequestsForContainer can be protected instead of public > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch, > Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, > Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, > Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14061320#comment-14061320 ] Mayank Bansal commented on YARN-1408: - [~sunilg], Can you check these test failures? Thanks, Mayank > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch, > Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, > Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, > Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14061101#comment-14061101
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12655590/Yarn-1408.11.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:red}-1 core tests{color}. The patch failed these unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager:
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServices
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesCapacitySched
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesAppsModification
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesNodes
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesApps
org.apache.hadoop.yarn.server.resourcemanager.webapp.TestRMWebServicesFairScheduler
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4296//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4296//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch,
> Yarn-1408.11.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch,
> Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch,
> Yarn-1408.9.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14060929#comment-14060929 ] Mayank Bansal commented on YARN-1408: - At [~jianhe] 's point . I think its good to check schedulerAttempt is not null before accessing it. Make Sense? Thanks, Mayank > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch, > Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, Yarn-1408.5.patch, > Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, Yarn-1408.9.patch, > Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14060699#comment-14060699
]
Wangda Tan commented on YARN-1408:
--
bq. is it possible that schedulerAttempt here is null? e.g. preemption happens
after the attempt completed.
After thinking about this, we can discuss following cases
1) New attempt started after old attempt completed
Scheduler (fair/capacity/fifo) will first set old attempt is stopped, but will
not remove it from existing SchedulerApplication. And recoverResourceRequests
will only take effect when attempt is not stopped
{code}
+if (!isStopped) {
+ appSchedulingInfo.updateResourceRequests(requests, true);
{code}
Then it will add a new attempt, but will use setCurrentAttempt method, will not
cause getCurrentAttempt become null.
And because the old attempt completed, we shouldn't recover new resource
request, so no-op is not a problem here too.
2) Application completed after old attempt completed
It is possible that SchedulerApplication become null here because old
application already completed and removed from scheduler. So add a null check
before
{code}
+schedulerAttempt.recoverResourceRequests(requests);
{code}
Should be enough.
Is it make sense to you, [~jianhe]/[~sunilg]?
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch,
> Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, Yarn-1408.5.patch,
> Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, Yarn-1408.9.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14060379#comment-14060379
]
Jian He commented on YARN-1408:
---
More comments after looking at the latest patch:
- is it possible that schedulerAttempt here is null? e.g. preemption happens
after the attempt completed.
{code}
SchedulerApplicationAttempt schedulerAttempt
= getCurrentAttemptForContainer(rmContainer.getContainerId());
schedulerAttempt.recoverResourceRequests(requests);
{code}
- AbstractYarnScheduler#recoverResourceRequest, how about renaming to
recoverResourceRequestForContainer ?
- assert the size of the requests. it can be empty and the assertion will be
skipped. similarly for CapacityScheduler test
{code}
List requests = rmContainer.getResourceRequests();
// Once recovered, resource request will be present again in app
for (ResourceRequest request : requests) {
Assert.assertEquals(1,
app.getResourceRequest(priority, request.getResourceName())
.getNumContainers());
}
{code}
- Alternatively, calling warnOrKillContainer twice and setting
WAIT_TIME_BEFORE_KILL to a small value may do the work.
{code}
// Create a preempt event by sending KILL event. In real cases,
// FairScheduler#warnOrKillContainer will perform below steps.
ContainerStatus status = SchedulerUtils.createPreemptedContainerStatus(
rmContainer.getContainerId(), SchedulerUtils.PREEMPTED_CONTAINER);
scheduler.recoverResourceRequest(rmContainer);
app.containerCompleted(rmContainer, status, RMContainerEventType.KILL);
{code}
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch,
> Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, Yarn-1408.5.patch,
> Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, Yarn-1408.9.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14060138#comment-14060138 ] Wangda Tan commented on YARN-1408: -- LGTM, +1 Thanks, > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch, > Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, Yarn-1408.5.patch, > Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, Yarn-1408.9.patch, > Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14059754#comment-14059754
]
Hadoop QA commented on YARN-1408:
-
{color:green}+1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12655374/Yarn-1408.10.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 core tests{color}. The patch passed unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager.
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4284//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4284//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch,
> Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, Yarn-1408.5.patch,
> Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, Yarn-1408.9.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14059637#comment-14059637
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12655357/Yarn-1408.10.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:red}-1 core tests{color}. The patch failed these unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager:
org.apache.hadoop.yarn.server.resourcemanager.applicationsmanager.TestAMRestart
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4282//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4282//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.10.patch,
> Yarn-1408.2.patch, Yarn-1408.3.patch, Yarn-1408.4.patch, Yarn-1408.5.patch,
> Yarn-1408.6.patch, Yarn-1408.7.patch, Yarn-1408.8.patch, Yarn-1408.9.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14059561#comment-14059561
]
Jian He commented on YARN-1408:
---
Thanks Sunil for the patch, some minor comments:
- add comments to recoverResourceRequest method about why doing this.
- we can use getCurrentAttemptForContainer here
{code}
ApplicationId appId = rmContainer.getContainerId()
.getApplicationAttemptId().getApplicationId();
SchedulerApplication schedulerApp = applications.get(appId);
SchedulerApplicationAttempt schedulerAttempt = schedulerApp
.getCurrentAppAttempt();
{code}
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.8.patch, Yarn-1408.9.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14059511#comment-14059511 ] Wangda Tan commented on YARN-1408: -- Thanks for update, will +1 if test failure is not relevant. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.8.patch, Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14059104#comment-14059104 ] Mayank Bansal commented on YARN-1408: - Thanks [~sunilg] for the patch. Patch Looks good, Can you check these teste failures. Thanks, Mayank > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.8.patch, Yarn-1408.9.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14058980#comment-14058980
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12655230/Yarn-1408.9.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:red}-1 core tests{color}. The patch failed these unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager:
org.apache.hadoop.yarn.server.resourcemanager.TestWorkPreservingRMRestart
org.apache.hadoop.yarn.server.resourcemanager.TestRMRestart
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4274//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4274//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.8.patch, Yarn-1408.9.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14058563#comment-14058563
]
Wangda Tan commented on YARN-1408:
--
Hi [~sunilg],
It seems your patch cannot apply against trunk, could you please rebase it?
I just reviewed your latest patch, some minor comments
1)
{code}
// If container state is moved to ACQUIRED, request will be empty.
if (requests == null || requests.isEmpty()) {
return;
}
{code}
Now request can only be null or non-empty, right?
2)
Could you update TestFairScheduler to wait for new container allocated at the
end of test as same as TestCapacityScheduler?
3)
It seems YARN-2181 changed RMContainerImpl and now
TestFairScheduler#testRecoverRequestAfterPreemption will encounter a NPE. You
can use "createSchedulingRequest" to avoid such error, like below
{code}
ApplicationAttemptId appAttemptId =
createSchedulingRequest(1 * 1024, "root.default", "user", 1, 1);
{code}
You can check TestFairScheduler#testChoiceOfPreemptedContainers as an example.
4)
Add timeout to tests you added.
Thanks,
Wangda
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.8.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14057764#comment-14057764
]
Hadoop QA commented on YARN-1408:
-
{color:green}+1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12655031/Yarn-1408.8.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 3 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 2.0.3) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 core tests{color}. The patch passed unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager.
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4260//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4260//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.8.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14057605#comment-14057605
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12655007/Yarn-1408.8.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 3 new
or modified test files.
{color:red}-1 javac{color:red}. The patch appears to cause the build to
fail.
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4259//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.8.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14057200#comment-14057200
]
Mayank Bansal commented on YARN-1408:
-
Thanks [~sunilg] for the patch.
Patch looks good , There are some minor comments
1. You current patch is not applying on the trunk, Please rebase on trunk.
2. There are lot of unwanted formatting changes, can you please revert them
back. Some examples are as follows
{code}
- .currentTimeMillis());
+.currentTimeMillis());
{code}
{code}
-RMContainer rmContainer =
-new RMContainerImpl(container, attemptId, node.getNodeID(),
- applications.get(attemptId.getApplicationId()).getUser(), rmContext,
- status.getCreationTime());
+RMContainer rmContainer = new RMContainerImpl(container, attemptId,
+node.getNodeID(), applications.get(attemptId.getApplicationId())
+.getUser(), rmContext, status.getCreationTime());
{code}
Please check this in all the patch.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14056482#comment-14056482
]
Jian He commented on YARN-1408:
---
ah, given this event is synchronously processed, the race I mentioned should
not happen, never mind.
{code}
rmContainer.handle(new RMContainerEvent(rmContainer.getContainerId(),
RMContainerEventType.ACQUIRED));
{code}
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14055925#comment-14055925 ] Wangda Tan commented on YARN-1408: -- But after think about this for a while, is it really a problem? Because container will be added to returnContainerList in this case, and AM should know that, right? [~jianhe]. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14055922#comment-14055922
]
Wangda Tan commented on YARN-1408:
--
Hi [~sunilg]
Thanks for find a new race condition :)
Actually, what [~jianhe] mentioned is not this one.
It is possible that, pullNewlyAllocatedContainer happened before container's
state transition to ACQUIRED because following code snippet
(SchedulerAppAttempt#pullNewlyAllocatedContainersAndNMTokens):
{code}
1 returnContainerList.add(container);
2 i.remove();
3 rmContainer.handle(new RMContainerEvent(rmContainer.getContainerId(),
RMContainerEventType.ACQUIRED));
{code}
It is possible that container killed after line-1 and before rmContainer handle
ACQUIRED event. And in such case, RR will be missed even if container preempted.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14055852#comment-14055852
]
Sunil G commented on YARN-1408:
---
Thanks [~jianhe] for the comment.
{code}
// Remove from the list of newly allocated containers if found
newlyAllocatedContainers.remove(rmContainer);
{code}
In *FicaSchedulerApp#containerCompleted*, the preempted container is removed
from newlyAllocatedContainers (as in patch). So
pullNewlyAllocatedContainersAndNMTokens will not get this container from
newlyAllocatedContainers. So duplicate containers will not reach AM. I missed
to add this to FSSchedulerApp. Please check this part and let me know if I am
inline with your scenario.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14055787#comment-14055787 ] Wangda Tan commented on YARN-1408: -- Thanks [~jianhe] for this commit, maybe a possible solution is move RMContainerImpl#setResourceRequest from AcquiredTransition to pullNewlyAllocatedContainer? > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14055753#comment-14055753 ] Jian He commented on YARN-1408: --- seems one race condition that if container is preempted before container moves to ACQUIRED state, but the container is already acquired by AM, the resource requests will be added back, and we'll allocate more containers than expected? > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14052239#comment-14052239 ] Wangda Tan commented on YARN-1408: -- Hi [~sunilg], Thanks for your reply, this make sense to me. This patch LGTM, +1 Could you take a look at this patch? [~vinodkv], [~mayank_bansal]. Thanks! > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch, > Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14052238#comment-14052238
]
Sunil G commented on YARN-1408:
---
Thank you [~leftnoteasy] for the comments.
bq.Can we put RMContainerImpl.setResourceRequests to RMContainer
As [~vinodkv] mentioned in YARN-2022, RMContainer usually is a read only
interface. So I feel we can place setResourceRequests in RMContainerImpl
itself and downcast the object as needed to invoke setResourceRequests.
bq.Can we verify after recovered, does RR in ASI contain node, rack, any
{code}
for (ResourceRequest request : requests) {
Assert.assertEquals(1,
app.getResourceRequest(priority, request.getResourceName())
.getNumContainers());
}
{code}
Here the loop is now running on “requests” and it has all node, rack, any
requests. This can check whether container count is 1 for that specific
request. So indirectly all 3 requests were verified as whether its present in
“app”. Please check this and let me know whether its inline with your thought.
Thank you.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14052119#comment-14052119
]
Hadoop QA commented on YARN-1408:
-
{color:green}+1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12654025/Yarn-1408.7.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 3 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 1.3.9) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 core tests{color}. The patch passed unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager.
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4197//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4197//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14052110#comment-14052110
]
Wangda Tan commented on YARN-1408:
--
[~sunilg],
Thanks update, just two minor comments,
1) Can we put RMContainerImpl.setResourceRequests to RMContainer to prevent
type cast such as
{code}
((RMContainerImpl)rmContainer).setResourceRequests(resourceRequestList);
{code}
2) For TestCapacityScheduler and TestFairScheduler,
Can we verify after recovered, does RR in ASI contain \{node, rack, any\}?
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.7.patch,
> Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14047755#comment-14047755
]
Wangda Tan commented on YARN-1408:
--
Hi [~sunilg],
Thanks for updating the patch, overall approach LGTM, some comments,
1)
bq. I think we can have a new api in appSchedulingInfo to return list of
ResourceRequests (node local, rack local and any).
I would suggest to modify existing appSchedulingInfo.allocate to return list of
RRs. There existed outstanding resource decrement logic in allocate(), we can
simply add decremented RR to a list and return them. It looks more like
by-product of ASI.allocate to me.
2)
{code}
if (type.equals(NodeType.NODE_LOCAL)) {
list.add(nodeRequests.get(hostName));
}
{code}
It's better to clone RR instead of add ref to list. It works, but it's better
to set a #container correctly and prevent RR changed in ASI in the future.
3) TestCapacityScheduler:
It's good to have a test for FairScheduler here too. I think we can put the
test to org.apache.hadoop.yarn.server.resourcemanager.scheduler, and make it
parameterized for Fair/Capacity/FIFO.
Two minor comment for TestCapacityScheduler.
3.1
{code}
for (ResourceRequest request : requests) {
// Skip the OffRack and RackLocal resource requests.
if (request.getResourceName().equals(node.getRackName())
|| request.getResourceName().equals(ResourceRequest.ANY)) {
Assert.assertEquals(request.getNumContainers(), 1);
continue;
}
// Resource request must have added back in RM after preempt event
handling.
Assert.assertNotNull(app.getResourceRequest(request.getPriority(),
request.getResourceName()));
}
{code}
We can make it simpler to,
{code}
for (ResourceRequest request : requests) {
// Resource request must have added back in RM after preempt event
handling.
Assert.assertEquals(1, app.getResourceRequest(request.getPriority(),
request.getResourceName()).getNumContainers());
}
{code}
Because we added them back, there's no difference between node/rack/any.
3.2
{code}
// allocate container
List containers = am1.allocate(new ArrayList(),
new ArrayList()).getAllocatedContainers();
{code}
Should we wait for containers allocated in a while loop? This works now because
previous we called "rm1.waitForState(nm1, ...)". But it's better to wait
container allocated explictly
Thanks,
Wangda
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14047694#comment-14047694
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12653155/Yarn-1408.6.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 1.3.9) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:red}-1 core tests{color}. The patch failed these unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager:
org.apache.hadoop.yarn.server.resourcemanager.ahs.TestRMApplicationHistoryWriter
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4145//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4145//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.6.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14046590#comment-14046590 ] Wangda Tan commented on YARN-1408: -- Hi [~sunilg], Thanks for working out this patch so fast! *A major problems I've seen.* ResourceRequest stored in RMContainerImpl should include rack/any RR, Currently, there's only one ResourceRequest stored in RMContainerImpl, this may should not enough for recovering in following cases: Case 1: RR may contain other fields like relaxLocality, etc. Assume a RR is node-local, the relaxLocaity=true (default), and it's rack-local/any RR's relaxLocality=false. In your current implementation, you cannot fully recover original RRs. Case 2: Rack-local RR will be missing. Assume a RR is node-local, when do resource allocation, the outstanding rack-local/any numContainer will be decreased, you can check AppSchedulingInfo#allocateNodeLocal for the logic of how outstanding rack/any #containers decreased. *My thoughts about how to implement this is:* In FiCaScheduler#allocate, appSchedulingInfo.allocate will be invoked. You can edit appSchedulingInfo.allocate to return a list a RRs, include node/rack/any if possible. Pass such RRs to RMContainerImpl And could you please elaborate on this? bq. AM would have asked for NodeLocal in another Hosts, which may not be able to recover. Does it make sense to you? I'll review minor issues and test cases in next cycle. Thanks, Wangda > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14046199#comment-14046199
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12652860/Yarn-1408.5.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 4 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 1.3.9) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:red}-1 core tests{color}. The patch failed these unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager:
org.apache.hadoop.yarn.server.resourcemanager.ahs.TestRMApplicationHistoryWriter
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/4122//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4122//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14045970#comment-14045970
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12652828/Yarn-1408.5.patch
against trunk revision .
{color:red}-1 patch{color}. The patch command could not apply the patch.
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/4119//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
>Assignee: Sunil G
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.5.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14045303#comment-14045303 ] Wangda Tan commented on YARN-1408: -- Thanks [~sunilg], looking forward your patch :) > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14044553#comment-14044553 ] Sunil G commented on YARN-1408: --- Hi [~leftnoteasy] As discussed in YARN-2143 last day, I am now using the AbstractScheduler api only to repopulate ResourceRequest to Scheduler. I am working on a patch based on this input and will upload a patch in a day. Thank you. :) > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14044548#comment-14044548 ] Wangda Tan commented on YARN-1408: -- Hi [~sunilg], do you have any updates on this issue, thanks :) > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14026243#comment-14026243 ] Sunil G commented on YARN-1408: --- Thank you Vinod. As discussed, I will work on a patch for storing ResourceRequest as transitory in RMContainer and restoring the same from killContainer API of AbstractYarnScheduler based on RMContainer state. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14026210#comment-14026210 ] Wangda Tan commented on YARN-1408: -- [~vinodkv], thanks for jumping in, +1 for put common killContainer API into AbstractYarnScheduler. I've filed YARN-2143 to track this effort. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14026192#comment-14026192 ] Vinod Kumar Vavilapalli commented on YARN-1408: --- bq. In this scenario, I hope we are trying to make this new killContainer api in AbstractYarnScheduler generic for these two schedulers for preemption Yup, let's do that. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14026187#comment-14026187 ] Sunil G commented on YARN-1408: --- Yes [~vinodkv]. +1 for the killContainer api in AbstractYarnScheduler. I have one doubt here, PreemptableResourceScheduler has killContainer api used in preemption context. FairScheduler is using warnOrKillContainer method to invoke KILL event on container in preemption context. In this scenario, I hope we are trying to make this new killContainer api in AbstractYarnScheduler generic for these two schedulers for preemption. Could you please share your thoughts. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14026159#comment-14026159 ] Vinod Kumar Vavilapalli commented on YARN-1408: --- Pitching in late, apologies. bq. After read code and thought about this problem, I think it's better to store ResourceRequests deducted when allocate a container: +1 for this proposal. It will add a transitory memory foot-print. Let's make sure we benchmark this. bq. 2. RMContainer can also store the AppSchedulingInfo object. This is leaking abstractions and is a slippery slope. It is kind of weird to have container talking to ASI. How about us adding a killContainer API in scheduler (AbstractYarnScheduler) which checks the container state and adds back the RR to ASI when needed? Killing a container is a generic functionality anyways. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14025274#comment-14025274 ] Wangda Tan commented on YARN-1408: -- [~sunilg], This approach looks good to me. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14021881#comment-14021881 ] Sunil G commented on YARN-1408: --- Thank you Wangda for the comment. I was thinking of using *AppSchedulingInfo* object itself. A reference of this object can be placed in RMContainer. Yes, there is a choice between *SchedulerApplicationAttempt* and *AppSchedulingInfo*. I feel based on heaviness, AppSchedulingInfo is better. AppSchedulingInfo#updateResourceRequests is synchronized also. If any other problems are there with AppSchedulingInfo, we can use SchedulerApplicationAttempt object. I will try summarizing the approach, please correct me if I am wrong. *Storage Part*: 1. RMContainer needs to store the informations such as priority, hostname, capability and relaxLocality to create a new ResourceRequest. 2. RMContainer can also store the AppSchedulingInfo object. From this object, AppSchedulingInfo#updateResourceRequests can be invoked later to add back ResourceRequest. *Transition Handling*: 1. RMContainer state transition: ALLOCATED to KILLED, a new transition named *PreemptTransition* can be invoked (extending FinishedTransition). 2. PreemptTransition can first create a new ResourceRequest instance with container count as 1 and with other details stored in RMContainer. * Invoke AppSchedulingInfo#updateResourceRequests with this new ResourceRequest instance * Invoke super.transition to handle the remaining FinishedTransition handling. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14020645#comment-14020645 ] Wangda Tan commented on YARN-1408: -- Thanks [~sunilg] for your comment, bq. May be here we can try add the required object to event, which can add back a ResurceRequest to scheduler with modified AppSchedulingInfo#updateResourceRequests. Is this what you also thought? I'm not sure what's the "required object" you mentioned. I think we don't need to link with respective scheduler. Maybe a SchedulerApplicationAttempt with updated AppSchedulingInfo#updateResourceRequests should be enough, do you think so? It seems to me that we don't need touch other specific scheduler implementations to me for this task. Wangda > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14019813#comment-14019813 ] Sunil G commented on YARN-1408: --- Thank you Wangda for the comment. bq.FairScheduler supports preemption as well, but it doesn't inherent PreemptableResourceScheduler interface. Yes. This will be cause similar problem in Fair also. I agree that we have to add back the ResourceRequest. Because as you mentioned in the example the user, which is AM here, will not get any update at all. In summary, the containers which are yet to be pulled off by AM, but killed by RM (preemption cases) are the problem. bq.If container transferred to ACQUIRED state, stored ResourceRequests will be removed directly. bq.If container killed/preempted before transferred to ACQUIRED state, we should recover stored ResourceRequests. Inline with this, if we want to recover ResourceRequest during RMContainer transitions(like KILL before ACQUIRED), we need a link back to the respective scheduler. May be here we can try add the required object to event, which can add back a ResurceRequest to scheduler with modified AppSchedulingInfo#updateResourceRequests. Is this what you also thought? > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14019798#comment-14019798 ] Wangda Tan commented on YARN-1408: -- [~sunilg], thanks for reply, bq. My doubt is in this context. I felt like we may not need a SchedulerApplicationAttempt in RMContainer itself. Rather I can do like keeping required information in RMContainer. So it will be more sepcific to the handling. bq. KILL event can happen not only from Preemption. I feel we need to deal issue with repect to preemption only, correct?. Other KILL event in RMContainer may be for a reserved container KILL etc. So we may need to do only in PreemptableResourceScheduler#killContainer. What do you feel? FairScheduler supports preemption as well, but it doesn't inherent PreemptableResourceScheduler interface. And do you agree that we need ALWAYS add Resource Request back when a container killed before AM acquired it (no matter this container is killed or not)? It's more like you've ordered something in Amazon, but you don't know Amazon already cancelled your order, it's not a good experience for user in this case. If you agree with me, it might be better to add logics in RMContainer transitions. :) bq. I agree with the concern of modifying definition, but to an extent I think it will be better if we can reuse with a recovery mode. +1 for extending and reusing existed interface. Wangda > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14019783#comment-14019783 ] Sunil G commented on YARN-1408: --- Hi [~leftnoteasy] bq.I meant we need keep a complete Resource Request, which includes Resource Request itself, rack/ANY level Resource Request, in case of the entire ResourceRequest removed when count reaches zero. I think we don't have difference here. , right? Yes. :) bq.pass a reference of SchedulerApplicationAttempt to RMContainer My doubt is in this context. I felt like we may not need a SchedulerApplicationAttempt in RMContainer itself. Rather I can do like keeping required information in RMContainer. So it will be more sepcific to the handling. bq.IMHO, it's better to put recover ResourceRequest logic in RMContainerImpl.FinishedTransition(), we can check if original state is Allocated and event is KILL. The benefit of this choice is we don't need separately modify FairScheduler and CapacityScheduler. Make sense? KILL event can happen not only from Preemption. I feel we need to deal issue with repect to preemption only, correct?. Other KILL event in RMContainer may be for a reserved container KILL etc. So we may need to do only in PreemptableResourceScheduler#killContainer. What do you feel? bq.Yes, but we cannot reuse this API without modifying its definition. The AppSchedulingInfo#updateResourceRequests will replace original ResourceRequests to new ResourceRequests, and our requirement is only increase the original ResourceRequest. I agree with the concern of modifying definition, but to an extent I think it will be better if we can reuse with a recovery mode. bq.And please note that, we need update QueueMetrics as well. +1 Yes. I agree. We have to do this also. Thank you. Sunil > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14019730#comment-14019730 ] Wangda Tan commented on YARN-1408: -- Hi [~sunilg], bq. I feel as per your design, we are keeping a much information than needed. Rather we can just keep only requried info for a ResourceRequest creation. I meant we need keep a complete Resource Request, which includes Resource Request itself, rack/ANY level Resource Request, in case of the entire ResourceRequest removed when count reaches zero. I think we don't have difference here. :), right? bq. I feel we can store enough information to create a new ResourceRequest at RMContainer itself. In FiCaSchedulerApp#allocate, a replica or partial information of ResourceRequest can be stored in RMContainer itself. +1 for this, since the resource requests is only a part of RMContainer, we don't need store it in app level bq. CapacityScheduler is extending PreemptableResourceScheduler interface and it has killContainer implementation. So preempt event handling can be done at killContainer. IMHO, it's better to put recover ResourceRequest logic in RMContainerImpl.FinishedTransition(), we can check if original state is Allocated and event is KILL. The benefit of this choice is we don't need separately modify FairScheduler and CapacityScheduler. Make sense? bq. api AppSchedulingInfo#updateResourceRequests can be used to update new ResourceRequest. Yes, but we cannot reuse this API without modifying its definition. The AppSchedulingInfo#updateResourceRequests will replace original ResourceRequests to new ResourceRequests, and our requirement is only increase the original ResourceRequest. We can either add a flag to this API indicate it's a recovering behavior or create a new API for it. And please note that, we need update QueueMetrics as well. Thanks, Wangda > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14019708#comment-14019708 ] Sunil G commented on YARN-1408: --- *Sorry I misunderstood, please discard my earlier comment*. I feel as per your design, we are keeping a much information than needed. Rather we can just keep only requried info for a ResourceRequest creation. bq.If container killed/preempted before transferred to ACQUIRED state, we should recover stored ResourceRequests. Container count is either decremented or that entry is removed from priority based map (AppSchedulingInfo), if the count is 0. So once removed, a complete Resource Request has to be added back with corresponding priority. CapacityScheduler is extending PreemptableResourceScheduler interface and it has killContainer implementation. So preempt event handling can be done at *killContainer*. I feel we can store enough information to create a new ResourceRequest at RMContainer itself. In FiCaSchedulerApp#allocate, a replica or partial information of ResourceRequest can be stored in RMContainer itself. And api AppSchedulingInfo#updateResourceRequests can be used to update new ResourceRequest. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14019696#comment-14019696 ] Sunil G commented on YARN-1408: --- bq. If container killed/preempted before transferred to ACQUIRED state, we should recover stored ResourceRequests. Container count is either decremented or that entry is removed from priority based map, if the count is 0. So once removed, a complete Resource Request has to be added back with corresponding priority. RMContainer has priority information and it can be used to get ResourceRequest with *AppSchedulingInfo#getResourceRequests* api. CapacityScheduler is extending PreemptableResourceScheduler interface and it has *killContainer* implementation. From this the corresponding FiCaSchedulerApp object can be found and thus can get the reference of AppSchedulingInfo. I am thinking this way the ResourceRequest can be recreated and added back if the Container is preempted at ACQUIRED/ALLOCATED etc. *killContainer* will be invoked at the preemption event and container state also can be found here. [~leftnoteasy], [~jianhe], [~mayank_bansal] is this approach fine? > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14019646#comment-14019646 ] Wangda Tan commented on YARN-1408: -- After read code and thought about this problem, I think it's better to store ResourceRequests deducted when allocate a container: * If container transferred to ACQUIRED state, stored ResourceRequests will be removed directly. * If container killed/preempted before transferred to ACQUIRED state, we should recover stored ResourceRequests. If we don't store ResourceRequests, it will be hard for us to re-create exactly same ResourceRequest from a container: container maybe allocated from a off-rack ResourceRequest, or ResourceRequest set some special fields like relaxLocality. A possible place to store ResourceRequest is SchedulerApplicationAttempt which covers Fair/CapacityScheduler together, and expose a new method like "recoverResourceRequest" and we can pass a reference of SchedulerApplicationAttempt to RMContainer. Once a container state become ACQUIRED, RMContainer can call this method to recover ResourceRequests. [~sunilg], [~jianhe], [~mayank_bansal], sounds like a doable plan? > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14008576#comment-14008576 ] Sunil G commented on YARN-1408: --- Yes. As I mentioned in earlier comment, it is better to add back/recreate the Resource Request back in CS. Also it will be better to do from CapacityScheduler#killContainer, as it is invoked directly in the preemption context. [~jianhe], [~mayank_bansal] please share your thoughts. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14007148#comment-14007148 ] Wangda Tan commented on YARN-1408: -- I think container should be preempt-able when its in allocate state, as race condition mentioned by [~jianhe], can we add a resource request to RMContainer? When allocate/kill within one AM heartbeat happened, we can add the resource request back. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14006908#comment-14006908 ] Sunil G commented on YARN-1408: --- bq. we may change it to decrement the resource request only when the container is pulled by the AM ? As [~jianhe] mentioned, this can create problem with subsequent NM heartbeats. Also I agree that the container in ALLOCATED state is the best place to do preemption, but this raise condition can come there. CapacityScheduler raises KILL event for RMContainer(for preemption). So a solution may be like recreate resource request back, if the RMContainer state is ALLOCATED/ACQUIRED here. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14006613#comment-14006613 ] Jian He commented on YARN-1408: --- Seems more problem with the approach I mentioned, if the request is not updated at the time container is allocated, and AM doesn't do the following allocate, more containers will be allocated from the same request when NM heartbeats > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14006548#comment-14006548 ] Mayank Bansal commented on YARN-1408: - I agree with [~jianhe] and [~devaraj.k] We should be able to preempt the container in ALLOCATED state. bq. oday the resource request is decremented when container is allocated. we may change it to decrement the resource request only when the container is pulled by the AM ? I am not sure if thats the right thing as you dont want to run into other race conditions when container is been allocated however capacity is given to some other AM's? > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14006465#comment-14006465 ] Jian He commented on YARN-1408: --- Hi [~sunilg], agree that we should remove container from newlyAllocatedContainers when preemption happens. As per the race condition you mentioned, we may also preempt ACQUIRED container? In fact, I think the best container to be preempted is the ALLOCATED container as these containers are not yet alive from the user's perspective. As per the race condition that [RM lost the resource request], today the resource request is decremented when container is allocated. we may change it to decrement the resource request only when the container is pulled by the AM ? We can do this separately if this makes sense. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Sub-task > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G >Assignee: Sunil G > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13994801#comment-13994801 ] Sunil G commented on YARN-1408: --- Please check the below scenario. After allocating a container to an application, CS will decrement its associated Resource Request info. Once this container is identified for preemption, preemption module in RM will do the container kill regardless whatever state the container is. I am assuming that state of one container is AQUIRED [Waiting for Launch event to become RUNNING]. And now this is marked for preemption, so container will get preempted. Hence Next heartbeat to AM has same container present in newlyAllocatedContainers and in completedContainers. [Allocation and Kill happened within an AM heartbeat cycle] An Invalid state transition [AQUIRED at KILLED] will be happened while processing from newlyAllocatedContainers in AM side. This will cause task to timeout after 30mins. If we try remove container from newlyAllocatedContainers, we can avoid invalid state transition. But this will cause task hang. [RM lost the resource request] As per initial explanation, RM has allocated a container and AM is waiting to get that container to assign for a task. Due to preemption, this has not been happened. Hence it will cause task to hang. I feel we can preempt those containers which are only in RUNNING state. [~devaraj.k] and [~curino], please share your thoughts. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Bug > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G > Fix For: 2.5.0 > > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13984152#comment-13984152
]
Devaraj K commented on YARN-1408:
-
Correction to the above comment.
Sorry for the delay Sunil.
{quote} So in some race conditions, it is possible that a container can get
KILLED by preemption even before it reach RUNNING state.
This scenario can be avoided if we can skip such containers which didnt reach
the RUNNING state during preemption.
May be in the following cycles this container will reach RUNNING state and the
can be considered for preemption.
{quote}
I think we don't need to wait for the container to move to RUNNING state for
preemption even if it is eligible. If the container is eligible for preemption,
the resources can be released with the current preemption cycle instead of
waiting for the next preemption cycle to change the container state to RUNNING,
so that it could save the wastage of the container launching and then killing.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Bug
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
> Fix For: 2.5.0
>
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13984151#comment-13984151 ] Devaraj K commented on YARN-1408: - bq. So in some race conditions, it is possible that a container can get KILLED by preemption even before it reach RUNNING state. This scenario can be avoided if we can skip such containers which didnt reach the RUNNING state during preemption. May be in the following cycles this container will reach RUNNING state and the can be considered for preemption. I think we don't need to wait for the container to move to RUNNING state for preemption even if it is eligible. If the container is eligible for preemption, the resources can be released with the current preemption cycle instead of waiting for the next preemption cycle to change the container state to RUNNING, so that it could save the wastage of the container launching and then killing. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Bug > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G > Fix For: 2.5.0 > > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, > Yarn-1408.4.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13919086#comment-13919086
]
Hadoop QA commented on YARN-1408:
-
{color:green}+1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12629000/Yarn-1408.4.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:green}+1 tests included{color}. The patch appears to include 1 new
or modified test files.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. There were no new javadoc warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 1.3.9) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 core tests{color}. The patch passed unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager.
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/3238//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/3238//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Bug
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
> Fix For: 2.4.0
>
> Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch,
> Yarn-1408.4.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.2#6252)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13901288#comment-13901288 ] Sunil G commented on YARN-1408: --- An update on this patch: For preemption policy, getLiveConatiners() is used to get the running containers. But container is added to this list when it is in ALLOCATED state itself. So in some race conditions, it is possible that a container can get KILLED by preemption even before it reach RUNNING state. This scenario can be avoided if we can skip such containers which didnt reach the RUNNING state during preemption. May be in the following cycles this container will reach RUNNING state and the can be considered for preemption. [~devaraj.k] please check this updated patch and let know your opinion. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Bug > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G > Fix For: 2.2.0 > > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.1.5#6160)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13859336#comment-13859336 ] Sunil G commented on YARN-1408: --- Hi Devaraj As per your comments, I have made the changes. 1. Need to handle the invalid transition and during the transition container to be removed from ContainerAllocationExpirer to avoid the timeout. [Sunil]: When we remove this extra preempted container from the newlyAllocatedContainers, the invalid transition got handled. Because, when heartbeat comes, this extra container will not be there in newlyAllocatedContainers and hence ACQUIRED event will not be fired at this container. 2. In the patch, trying to remove from newlyAllocatedContainers. This can be removed directly from newlyAllocatedContainers using java.util.List.remove(Object o), instead of iterating, checking and then removing. [Sunil]: Yes, i changed it by removing directly from the list 3. Can you also add test to demonstrate this case. [Sunil]: Change has done to remove an element from the newlyAllocatedContainers. There are no functions added. Now the verification is done by manual testing to ensure the removal is performed. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Bug > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G > Fix For: 2.2.0 > > Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.1.5#6160)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[ https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13830993#comment-13830993 ] Sunil G commented on YARN-1408: --- Thank you Devaraj for reviewing the patch. I will change and update the patch as per the comments. > Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task > timeout for 30mins > -- > > Key: YARN-1408 > URL: https://issues.apache.org/jira/browse/YARN-1408 > Project: Hadoop YARN > Issue Type: Bug > Components: resourcemanager >Affects Versions: 2.2.0 >Reporter: Sunil G > Fix For: 2.2.0 > > Attachments: Yarn-1408.1.patch, Yarn-1408.patch > > > Capacity preemption is enabled as follows. > * yarn.resourcemanager.scheduler.monitor.enable= true , > * > yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy > Queue = a,b > Capacity of Queue A = 80% > Capacity of Queue B = 20% > Step 1: Assign a big jobA on queue a which uses full cluster capacity > Step 2: Submitted a jobB to queue b which would use less than 20% of cluster > capacity > JobA task which uses queue b capcity is been preempted and killed. > This caused below problem: > 1. New Container has got allocated for jobA in Queue A as per node update > from an NM. > 2. This container has been preempted immediately as per preemption. > Here ACQUIRED at KILLED Invalid State exception came when the next AM > heartbeat reached RM. > ERROR > org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: > Can't handle this event at current state > org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: > ACQUIRED at KILLED > This also caused the Task to go for a timeout for 30minutes as this Container > was already killed by preemption. > attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13821193#comment-13821193
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12613569/Yarn-1408.1.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author
tags.
{color:red}-1 tests included{color}. The patch doesn't appear to include
any new or modified tests.
Please justify why no new tests are needed for this
patch.
Also please list what manual steps were performed to
verify this patch.
{color:green}+1 javac{color}. The applied patch does not increase the
total number of javac compiler warnings.
{color:green}+1 javadoc{color}. The javadoc tool did not generate any
warning messages.
{color:green}+1 eclipse:eclipse{color}. The patch built with
eclipse:eclipse.
{color:green}+1 findbugs{color}. The patch does not introduce any new
Findbugs (version 1.3.9) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase
the total number of release audit warnings.
{color:green}+1 core tests{color}. The patch passed unit tests in
hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager.
{color:green}+1 contrib tests{color}. The patch passed contrib unit tests.
Test results:
https://builds.apache.org/job/PreCommit-YARN-Build/2444//testReport/
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/2444//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Bug
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
> Fix For: 2.2.0
>
> Attachments: Yarn-1408.1.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.1#6144)
[jira] [Commented] (YARN-1408) Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task timeout for 30mins
[
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13821108#comment-13821108
]
Hadoop QA commented on YARN-1408:
-
{color:red}-1 overall{color}. Here are the results of testing the latest
attachment
http://issues.apache.org/jira/secure/attachment/12613538/Yarn-1408.patch
against trunk revision .
{color:red}-1 patch{color}. The patch command could not apply the patch.
Console output: https://builds.apache.org/job/PreCommit-YARN-Build/2440//console
This message is automatically generated.
> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task
> timeout for 30mins
> --
>
> Key: YARN-1408
> URL: https://issues.apache.org/jira/browse/YARN-1408
> Project: Hadoop YARN
> Issue Type: Bug
> Components: resourcemanager
>Affects Versions: 2.2.0
>Reporter: Sunil G
> Fix For: 2.2.0
>
> Attachments: Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
> * yarn.resourcemanager.scheduler.monitor.enable= true ,
> *
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b which would use less than 20% of cluster
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM
> heartbeat reached RM.
> ERROR
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl:
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container
> was already killed by preemption.
> attempt_1380289782418_0003_m_00_0 Timed out after 1800 secs
--
This message was sent by Atlassian JIRA
(v6.1#6144)
