subject:"\[jira\] \[Commented\] \(HBASE\-9514\) Prevent region from assigning before log splitting is done"

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Hudson (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783598#comment-13783598
 ] 

Hudson commented on HBASE-9514:
---

SUCCESS: Integrated in hbase-0.96 #114 (See 
[https://builds.apache.org/job/hbase-0.96/114/])
HBASE-9514 Prevent region from assigning before log splitting is done - 
ADDENDUM (jxiang: rev 1528247)
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/util/HBaseFsck.java


> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.98.0, 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Hudson (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783551#comment-13783551
 ] 

Hudson commented on HBASE-9514:
---

SUCCESS: Integrated in HBase-TRUNK-on-Hadoop-2.0.0 #771 (See 
[https://builds.apache.org/job/HBase-TRUNK-on-Hadoop-2.0.0/771/])
HBASE-9514 Prevent region from assigning before log splitting is done (jxiang: 
rev 1528227)
* 
/hbase/trunk/hbase-client/src/main/java/org/apache/hadoop/hbase/ClusterStatus.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/IntegrationTestBase.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/AddColumnAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/CompactRandomRegionOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/DumpClusterStatusAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/FlushRandomRegionOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MergeRandomAdjacentRegionsOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MoveRandomRegionOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MoveRegionsOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/factories/SlowDeterministicMonkeyFactory.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/test/IntegrationTestBigLinkedList.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/AssignmentManager.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/CatalogJanitor.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/DeadServer.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/GeneralBulkAssigner.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/ServerManager.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/ClosedRegionHandler.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/EnableTableHandler.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/MetaServerShutdownHandler.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/ServerShutdownHandler.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/util/HBaseFsck.java
* 
/hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/TestDrainingServer.java
* 
/hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestAssignmentManagerOnCluster.java
* 
/hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestSplitTransactionOnCluster.java


> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.98.0, 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Hudson (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783535#comment-13783535
 ] 

Hudson commented on HBASE-9514:
---

FAILURE: Integrated in HBase-TRUNK #4581 (See 
[https://builds.apache.org/job/HBase-TRUNK/4581/])
HBASE-9514 Prevent region from assigning before log splitting is done (jxiang: 
rev 1528227)
* 
/hbase/trunk/hbase-client/src/main/java/org/apache/hadoop/hbase/ClusterStatus.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/IntegrationTestBase.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/AddColumnAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/CompactRandomRegionOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/DumpClusterStatusAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/FlushRandomRegionOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MergeRandomAdjacentRegionsOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MoveRandomRegionOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MoveRegionsOfTableAction.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/factories/SlowDeterministicMonkeyFactory.java
* 
/hbase/trunk/hbase-it/src/test/java/org/apache/hadoop/hbase/test/IntegrationTestBigLinkedList.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/AssignmentManager.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/CatalogJanitor.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/DeadServer.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/GeneralBulkAssigner.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/ServerManager.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/ClosedRegionHandler.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/EnableTableHandler.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/MetaServerShutdownHandler.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/ServerShutdownHandler.java
* 
/hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/util/HBaseFsck.java
* 
/hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/TestDrainingServer.java
* 
/hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestAssignmentManagerOnCluster.java
* 
/hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestSplitTransactionOnCluster.java


> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.98.0, 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Hudson (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783536#comment-13783536
 ] 

Hudson commented on HBASE-9514:
---

FAILURE: Integrated in hbase-0.96-hadoop2 #68 (See 
[https://builds.apache.org/job/hbase-0.96-hadoop2/68/])
HBASE-9514 Prevent region from assigning before log splitting is done - 
ADDENDUM (jxiang: rev 1528247)
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/util/HBaseFsck.java
HBASE-9514 Prevent region from assigning before log splitting is done (jxiang: 
rev 1528231)
* 
/hbase/branches/0.96/hbase-client/src/main/java/org/apache/hadoop/hbase/ClusterStatus.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/IntegrationTestBase.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/AddColumnAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/CompactRandomRegionOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/DumpClusterStatusAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/FlushRandomRegionOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MergeRandomAdjacentRegionsOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MoveRandomRegionOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MoveRegionsOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/factories/SlowDeterministicMonkeyFactory.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/test/IntegrationTestBigLinkedList.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/AssignmentManager.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/CatalogJanitor.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/DeadServer.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/GeneralBulkAssigner.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/ServerManager.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/ClosedRegionHandler.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/EnableTableHandler.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/MetaServerShutdownHandler.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/ServerShutdownHandler.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/util/HBaseFsck.java
* 
/hbase/branches/0.96/hbase-server/src/test/java/org/apache/hadoop/hbase/TestDrainingServer.java
* 
/hbase/branches/0.96/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestAssignmentManagerOnCluster.java
* 
/hbase/branches/0.96/hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestSplitTransactionOnCluster.java


> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.98.0, 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Hudson (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783521#comment-13783521
 ] 

Hudson commented on HBASE-9514:
---

FAILURE: Integrated in hbase-0.96 #113 (See 
[https://builds.apache.org/job/hbase-0.96/113/])
HBASE-9514 Prevent region from assigning before log splitting is done (jxiang: 
rev 1528231)
* 
/hbase/branches/0.96/hbase-client/src/main/java/org/apache/hadoop/hbase/ClusterStatus.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/IntegrationTestBase.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/AddColumnAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/CompactRandomRegionOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/DumpClusterStatusAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/FlushRandomRegionOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MergeRandomAdjacentRegionsOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MoveRandomRegionOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/actions/MoveRegionsOfTableAction.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/chaos/factories/SlowDeterministicMonkeyFactory.java
* 
/hbase/branches/0.96/hbase-it/src/test/java/org/apache/hadoop/hbase/test/IntegrationTestBigLinkedList.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/AssignmentManager.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/CatalogJanitor.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/DeadServer.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/GeneralBulkAssigner.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/RegionStates.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/ServerManager.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/ClosedRegionHandler.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/EnableTableHandler.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/MetaServerShutdownHandler.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/master/handler/ServerShutdownHandler.java
* 
/hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/util/HBaseFsck.java
* 
/hbase/branches/0.96/hbase-server/src/test/java/org/apache/hadoop/hbase/TestDrainingServer.java
* 
/hbase/branches/0.96/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestAssignmentManagerOnCluster.java
* 
/hbase/branches/0.96/hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestSplitTransactionOnCluster.java


> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.98.0, 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783440#comment-13783440
 ] 

Jimmy Xiang commented on HBASE-9514:


Fixed. Thanks JD and Enis.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.98.0, 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783432#comment-13783432
 ] 

Jimmy Xiang commented on HBASE-9514:


JD: HBaseFsck#setDisplayFullReport is a static method in trunk, but not in 
0.96.  Let me fix it.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.98.0, 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783429#comment-13783429
 ] 

Enis Soztutar commented on HBASE-9514:
--

yep, that is the compilation error i mentioned above. Jimmy, can you please do 
an addendum. 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.98.0, 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Jean-Daniel Cryans (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783425#comment-13783425
 ] 

Jean-Daniel Cryans commented on HBASE-9514:
---

Just refreshed my repo, did a mvn clean and now I see:

{quote}
[ERROR] Failed to execute goal 
org.apache.maven.plugins:maven-compiler-plugin:2.5.1:testCompile 
(default-testCompile) on project hbase-it: Compilation failure
[ERROR] 
/Users/jdcryans/svn/0.96-commit/hbase-it/src/test/java/org/apache/hadoop/hbase/test/IntegrationTestBigLinkedList.java:[737,19]
 non-static method setDisplayFullReport() cannot be referenced from a static 
context
{quote}

Could this come from this jira?

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.98.0, 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783400#comment-13783400
 ] 

Jimmy Xiang commented on HBASE-9514:


This issue is that this kind of test takes a while to run. We'd like to collect 
as much information as possible if it fails so that we can triage any failure 
quick.
Thanks for the review.  I am going to integrated the patch into trunk and 0.96. 
 We can address remaining issues in separate jiras.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783336#comment-13783336
 ] 

Enis Soztutar commented on HBASE-9514:
--

Sorry, i still did not get why we are running verify 2 or 3 times after it 
failed. If test fails, you can manually run verify again on the table data, 
since it won't be deleted. Regardless, let's get this version in and get RC 
rolling I think. 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Hadoop QA (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783311#comment-13783311
 ] 

Hadoop QA commented on HBASE-9514:
--

{color:green}+1 overall{color}.  Here are the results of testing the latest 
attachment 
  http://issues.apache.org/jira/secure/attachment/12606160/trunk-9514_v5.5.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 38 new 
or modified tests.

{color:green}+1 hadoop1.0{color}.  The patch compiles against the hadoop 
1.0 profile.

{color:green}+1 hadoop2.0{color}.  The patch compiles against the hadoop 
2.0 profile.

{color:green}+1 javadoc{color}.  The javadoc tool did not generate any 
warning messages.

{color:green}+1 javac{color}.  The applied patch does not increase the 
total number of javac compiler warnings.

{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 lineLengths{color}.  The patch does not introduce lines 
longer than 100

  {color:green}+1 site{color}.  The mvn site goal succeeds with this patch.

{color:green}+1 core tests{color}.  The patch passed unit tests in .

Test results: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//testReport/
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-protocol.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-thrift.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-client.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-examples.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-prefix-tree.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7433//console

This message is automatically generated.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783223#comment-13783223
 ] 

Jimmy Xiang commented on HBASE-9514:


The javadoc issue is not caused by this patch. Totally, there are not many 
javadoc warnings. We can fix them all in a separate issue.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.5.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783203#comment-13783203
 ] 

Enis Soztutar commented on HBASE-9514:
--

Left some comments at RB. 
I am +1, except for the changes in ITBLL. I don't think we should mask verify 
failures in any way. 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Devaraj Das (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783189#comment-13783189
 ] 

Devaraj Das commented on HBASE-9514:


bq. The test succeeded yesterday for 1B records in total.
Excellent! Let's get it in.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread stack (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783187#comment-13783187
 ] 

stack commented on HBASE-9514:
--

You +1 on it [~enis]?

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783180#comment-13783180
 ] 

Enis Soztutar commented on HBASE-9514:
--

The test succeeded yesterday for 1B records in total. 
{code}
13/10/01 04:44:57 INFO test.IntegrationTestBigLinkedList$Loop: Verify finished 
with succees. Total nodes=10
{code}



> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-10-01 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13783173#comment-13783173
 ] 

Jimmy Xiang commented on HBASE-9514:


I can fix the javadoc complaint and the test failure.  We can get this in after 
[~enis] reviews it.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-30 Thread stack (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13782659#comment-13782659
 ] 

stack commented on HBASE-9514:
--

Can you repeat the above fail Jimmy?  Does the javadoc complaint come in w/ 
this patch?  Want me to review it again?

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-30 Thread Hadoop QA (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13782619#comment-13782619
 ] 

Hadoop QA commented on HBASE-9514:
--

{color:red}-1 overall{color}.  Here are the results of testing the latest 
attachment 
  http://issues.apache.org/jira/secure/attachment/12606036/trunk-9514_v5.4.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 38 new 
or modified tests.

{color:green}+1 hadoop1.0{color}.  The patch compiles against the hadoop 
1.0 profile.

{color:green}+1 hadoop2.0{color}.  The patch compiles against the hadoop 
2.0 profile.

{color:red}-1 javadoc{color}.  The javadoc tool appears to have generated 1 
warning messages.

{color:green}+1 javac{color}.  The applied patch does not increase the 
total number of javac compiler warnings.

{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 lineLengths{color}.  The patch does not introduce lines 
longer than 100

  {color:green}+1 site{color}.  The mvn site goal succeeds with this patch.

 {color:red}-1 core tests{color}.  The patch failed these unit tests:
   org.apache.hadoop.hbase.master.TestMasterFailover

Test results: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//testReport/
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-protocol.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-thrift.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-client.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-examples.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-prefix-tree.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7425//console

This message is automatically generated.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.4.patch, 
> trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-30 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13782561#comment-13782561
 ] 

Enis Soztutar commented on HBASE-9514:
--

v5.3 has a simple compilation error. Do you mind putting this version to RB as 
well. 
{code}
+if (!runVerify(outputDir, numReducers, expectedNumNodes, true)) {
+  LOG.info("Verify.verify failed, let's stop CM and verify again");
+  it.cleanUpMonkey("Monkey broke something");
+  if (!runVerify(outputDir, numReducers, expectedNumNodes, false)) {
+throw new RuntimeException("Verify.verify failed");
+  }
{code}
I did not get why we are running verify again. No matter what, if verify failed 
once, we should fail the test. 

Other changes look ok to me. although, it is still hard to follow what the edge 
cases might be. The tests are still running. 



> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-30 Thread Hadoop QA (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13782424#comment-13782424
 ] 

Hadoop QA commented on HBASE-9514:
--

{color:red}-1 overall{color}.  Here are the results of testing the latest 
attachment 
  http://issues.apache.org/jira/secure/attachment/12605998/trunk-9514_v5.3.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 38 new 
or modified tests.

{color:red}-1 hadoop1.0{color}.  The patch failed to compile against the 
hadoop 1.0 profile.

Console output: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7423//console

This message is automatically generated.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.3.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-30 Thread Hadoop QA (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13782073#comment-13782073
 ] 

Hadoop QA commented on HBASE-9514:
--

{color:green}+1 overall{color}.  Here are the results of testing the latest 
attachment 
  http://issues.apache.org/jira/secure/attachment/12605926/trunk-9514_v5.2.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 27 new 
or modified tests.

{color:green}+1 hadoop1.0{color}.  The patch compiles against the hadoop 
1.0 profile.

{color:green}+1 hadoop2.0{color}.  The patch compiles against the hadoop 
2.0 profile.

{color:green}+1 javadoc{color}.  The javadoc tool did not generate any 
warning messages.

{color:green}+1 javac{color}.  The applied patch does not increase the 
total number of javac compiler warnings.

{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 lineLengths{color}.  The patch does not introduce lines 
longer than 100

  {color:green}+1 site{color}.  The mvn site goal succeeds with this patch.

{color:green}+1 core tests{color}.  The patch passed unit tests in .

Test results: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//testReport/
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-prefix-tree.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-client.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-protocol.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-examples.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-thrift.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7419//console

This message is automatically generated.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-30 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13782065#comment-13782065
 ] 

Enis Soztutar commented on HBASE-9514:
--

Running the test with v5.2. Will report back with results. 
{code}
hbase org.apache.hadoop.hbase.test.IntegrationTestBigLinkedList --monkey 
slowDeterministic Loop 5 8 2500 /user/enis/IntegrationTestBigLinkedList 8 
>/tmp/biglinkedlist.log 2>&1
{code}


> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.2.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-29 Thread Hadoop QA (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13781574#comment-13781574
 ] 

Hadoop QA commented on HBASE-9514:
--

{color:green}+1 overall{color}.  Here are the results of testing the latest 
attachment 
  http://issues.apache.org/jira/secure/attachment/12605845/trunk-9514_v5.1.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 38 new 
or modified tests.

{color:green}+1 hadoop1.0{color}.  The patch compiles against the hadoop 
1.0 profile.

{color:green}+1 hadoop2.0{color}.  The patch compiles against the hadoop 
2.0 profile.

{color:green}+1 javadoc{color}.  The javadoc tool did not generate any 
warning messages.

{color:green}+1 javac{color}.  The applied patch does not increase the 
total number of javac compiler warnings.

{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 lineLengths{color}.  The patch does not introduce lines 
longer than 100

  {color:green}+1 site{color}.  The mvn site goal succeeds with this patch.

{color:green}+1 core tests{color}.  The patch passed unit tests in .

Test results: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//testReport/
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-protocol.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-thrift.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-client.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-examples.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-prefix-tree.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7416//console

This message is automatically generated.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.1.patch, 
> trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-29 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13781551#comment-13781551
 ] 

Jimmy Xiang commented on HBASE-9514:


Still not sure about why it fails even if there is no data loss.  Yes, please 
review it.  We can commit it and keep looking into it. Thanks.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-29 Thread stack (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13781526#comment-13781526
 ] 

stack commented on HBASE-9514:
--

Good stuff Jimmy.  So, test fails because we don't recover fast enough or 
something?  Should we commit what you have?   I can do a pass over it no 
problem.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.1.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-29 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13781502#comment-13781502
 ] 

Jimmy Xiang commented on HBASE-9514:


Here is the list of changes:
1. fixed a bug in AM#assign(line ~2645), when bulk assign fails, each region 
should be assigned again, otherwise, they will be stuck in transition;
2. fixed a bug in AM#unassign(line ~2461), if region is offline, assign it 
again (moved to final block, so all scenarios are covered);
3. in RegionStates if the last hosting region server is online, get the 
server's info to confirm it has the expected start code (may be too 
conservative, hasn't seen it in my test yet);
4. in AM, force region state offline, if force new plan, check meta to make 
sure the last assignment is not changed (may be too conservative, hasn't seen 
it in my test yet);
5. enhanced bulk assign a little so that if a region is already assign, no need 
to force assign.

I have a new patch in testing now. The new patch has the following changes:
1. added a CM action to log cluster status every 90 seconds so we know details 
about regions in transition;
2. added a hbck check after verification failure so that we know if the cluster 
is consistent, i.e., any region is lost/unassigned;
3. added another verify with CM disabled after verification failure so we know 
if we really have data loss.

It seems that there is no data loss now since 3. shows ok while the test still 
fails.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-29 Thread stack (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13781474#comment-13781474
 ] 

stack commented on HBASE-9514:
--

What are the main changes in this patch from what you posted previously Jimmy 
(seems reasonable after skimming first 1/5th but I have reviewed a bunch of 
this already).

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-27 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13780268#comment-13780268
 ] 

Jimmy Xiang commented on HBASE-9514:


I may dump some cluster info periodically if possible.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-27 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13780262#comment-13780262
 ] 

Jimmy Xiang commented on HBASE-9514:


v5 is what I am testing now. I am going to add some hbck checking to the end of 
IntegrationTestBigLinkedList if the test fails. So that we can see if any 
region is stuck in transition, or unassigned at all.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-27 Thread Hadoop QA (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13780240#comment-13780240
 ] 

Hadoop QA commented on HBASE-9514:
--

{color:green}+1 overall{color}.  Here are the results of testing the latest 
attachment 
  http://issues.apache.org/jira/secure/attachment/12605481/trunk-9514_v5.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 30 new 
or modified tests.

{color:green}+1 hadoop1.0{color}.  The patch compiles against the hadoop 
1.0 profile.

{color:green}+1 hadoop2.0{color}.  The patch compiles against the hadoop 
2.0 profile.

{color:green}+1 javadoc{color}.  The javadoc tool did not generate any 
warning messages.

{color:green}+1 javac{color}.  The applied patch does not increase the 
total number of javac compiler warnings.

{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 lineLengths{color}.  The patch does not introduce lines 
longer than 100

  {color:green}+1 site{color}.  The mvn site goal succeeds with this patch.

{color:green}+1 core tests{color}.  The patch passed unit tests in .

Test results: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//testReport/
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-protocol.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-thrift.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-client.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-examples.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-prefix-tree.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7405//console

This message is automatically generated.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Fix For: 0.96.0
>
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch, trunk-9514_v5.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-26 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13779158#comment-13779158
 ] 

Jimmy Xiang commented on HBASE-9514:


[~jeffreyz], yes, it should cover this scenario as well.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-26 Thread Jeffrey Zhong (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13779146#comment-13779146
 ] 

Jeffrey Zhong commented on HBASE-9514:
--

[~jxiang] Will your patch cover the scenario HBASE-9665 where balancer starts 
to move a region while the server hosting the regions dies then the region got 
lost due to ZK RIT state is messed up with the two concurrent assignments?

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-23 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13775654#comment-13775654
 ] 

Jimmy Xiang commented on HBASE-9514:


Almost there. Without schema change/master restart actions, I got it green.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-23 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13775634#comment-13775634
 ] 

Enis Soztutar commented on HBASE-9514:
--

Any update on this [~jxiang]? 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-19 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13772170#comment-13772170
 ] 

Jimmy Xiang commented on HBASE-9514:


I have fixed that one. I am fixing holes related to split/merge now.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-19 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13772150#comment-13772150
 ] 

Enis Soztutar commented on HBASE-9514:
--

Thanks Jimmy. The test I ran failed at the generator this time, because of two 
regions was stuck on FAILED_CLOSE:
{code}
4d9c5d2fdcad2855bd7100932734bede
IntegrationTestBigLinkedList,\x0B\xFF\x06\x80\xA5\xF2\x0F\xE0\xCD\x9CDh\xA1\xD8<\xDA,1379559344242.4d9c5d2fdcad2855bd7100932734bede.
 state=FAILED_CLOSE, ts=Thu Sep 19 03:47:32 UTC 2013 (53154s ago), 
server=hor8n01,60020,1379562089022   53154621 
{code}
This is definitely related to the latest patch. I can dig a bit more. 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-18 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13771545#comment-13771545
 ] 

Jimmy Xiang commented on HBASE-9514:


[~enis], there are still some issues with it (v4).  My test is still failing. I 
will let you know after I get a patch passing the test. Thanks.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-18 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13771541#comment-13771541
 ] 

Enis Soztutar commented on HBASE-9514:
--

The patch at RB (v4) looks mostly ok to me, although I must admit that I do not 
get the full details of AM internals. I am running the ITBL to see whether it 
fixes that problem. 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-17 Thread stack (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13770269#comment-13770269
 ] 

stack commented on HBASE-9514:
--

I'm trying to fix TestHLog.  Its broke.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-17 Thread Hadoop QA (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13770250#comment-13770250
 ] 

Hadoop QA commented on HBASE-9514:
--

{color:red}-1 overall{color}.  Here are the results of testing the latest 
attachment 
  http://issues.apache.org/jira/secure/attachment/12603700/trunk-9514_v3.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 18 new 
or modified tests.

{color:green}+1 hadoop1.0{color}.  The patch compiles against the hadoop 
1.0 profile.

{color:green}+1 hadoop2.0{color}.  The patch compiles against the hadoop 
2.0 profile.

{color:green}+1 javadoc{color}.  The javadoc tool did not generate any 
warning messages.

{color:green}+1 javac{color}.  The applied patch does not increase the 
total number of javac compiler warnings.

{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 lineLengths{color}.  The patch does not introduce lines 
longer than 100

  {color:green}+1 site{color}.  The mvn site goal succeeds with this patch.

 {color:red}-1 core tests{color}.  The patch failed these unit tests:
   org.apache.hadoop.hbase.regionserver.wal.TestHLog

Test results: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//testReport/
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-protocol.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-client.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-examples.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-prefix-tree.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7280//console

This message is automatically generated.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-17 Thread stack (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13770179#comment-13770179
 ] 

stack commented on HBASE-9514:
--

[~eclark] Doesn't this patch already take locks per region to avoid clash of am 
and ssh?

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-17 Thread Hadoop QA (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13770152#comment-13770152
 ] 

Hadoop QA commented on HBASE-9514:
--

{color:red}-1 overall{color}.  Here are the results of testing the latest 
attachment 
  http://issues.apache.org/jira/secure/attachment/12603655/trunk-9514_v2.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 18 new 
or modified tests.

{color:green}+1 hadoop1.0{color}.  The patch compiles against the hadoop 
1.0 profile.

{color:green}+1 hadoop2.0{color}.  The patch compiles against the hadoop 
2.0 profile.

{color:green}+1 javadoc{color}.  The javadoc tool did not generate any 
warning messages.

{color:green}+1 javac{color}.  The applied patch does not increase the 
total number of javac compiler warnings.

{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 lineLengths{color}.  The patch does not introduce lines 
longer than 100

  {color:green}+1 site{color}.  The mvn site goal succeeds with this patch.

 {color:red}-1 core tests{color}.  The patch failed these unit tests:
   org.apache.hadoop.hbase.regionserver.wal.TestHLog

Test results: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//testReport/
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-prefix-tree.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-client.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-protocol.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-examples.html
Findbugs warnings: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: 
https://builds.apache.org/job/PreCommit-HBASE-Build/7275//console

This message is automatically generated.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-17 Thread Elliott Clark (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13770126#comment-13770126
 ] 

Elliott Clark commented on HBASE-9514:
--

So I chatted with [~jxiang] about this patch.  For me it seems like the region 
locks are the perfect solution here.  Have the master grab region locks when 
the regionserver goes down.  Then the splitting server can grab locks.

That should allow the master to ride over any deaths pretty easily, and it uses 
the already in hbase mechanisms.  However it's a larger patch than the one here.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch, trunk-9514_v2.patch, 
> trunk-9514_v3.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-17 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13769768#comment-13769768
 ] 

Jimmy Xiang commented on HBASE-9514:


That's a good point.  Let me fix it.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-17 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13769660#comment-13769660
 ] 

Jimmy Xiang commented on HBASE-9514:


bq. should we rename AM.acquireLock -> AM.acquireRegionLock()
Sure, will do.
bq. Why not do this for meta?
Let me think how to cover meta as well.
bq. Is it safe to expire a server like this. It means the master cannot connect 
to it, but it may still have the zk lease. 
We have waited for longer than the ZK session timeout. It should be expired.  
It is ok to expire twice.

The timeout should be active. I will fix it.

The idea of the patch is to remember the last known region server a region is 
assigned to. Whenever we try to assign a region, we check if the last known 
region server of the region is done with log splitting. If not, we don't assign 
it, and let SSH to complete log splitting and re-assign. We clear the last 
known region server info when SSH finishes log splitting, or the region is 
properly closed. The idea is simple but there are several racing points to take 
care.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-17 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13769761#comment-13769761
 ] 

Enis Soztutar commented on HBASE-9514:
--

bq. We have waited for longer than the ZK session timeout. It should be 
expired. It is ok to expire twice.
My understanding is that server expiry should ONLY come from a zookeeper 
session timeout. The master being not able to connect to RS for more than zk 
session timeout does not necessarily mean that the session has actually 
expired. If a network partition happens and master cannot talk to RS, but RS 
still holds the zk lease, then the master will think that the server is dead, 
while RS will happily continue to serve the region. Given that the RS will be 
getting YouAreDeadException if it talks the master afterwards, and we are 
forcing lease recovery on the RS logs on log splitting, but I fear, while this 
process is going on there will be an inconsistency window where master will 
think RS is dead, while it may not be. 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-16 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13769237#comment-13769237
 ] 

Enis Soztutar commented on HBASE-9514:
--

A couple of comments:
 - should we rename AM.acquireLock -> AM.acquireRegionLock() 
 - Why not do this for meta? 
{code}
+if (!region.isMetaRegion() &&
+regionStates.wasRegionOnDeadServer(encodedName)) {
{code}

Is it safe to expire a server like this. It means the master cannot connect to 
it, but it may still have the zk lease. 
{code}
+} else {
+  LOG.info(server + " is not reachable, expire it");
+  serverManager.expireServer(server);
+}
{code}

- Should we rename RegionStates.logSplit() -> markAssignable() or something 
like it. 

Is this timeout intended to be active?
{code}
+  @Test //(timeout=6)
{code}

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-16 Thread stack (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13768966#comment-13768966
 ] 

stack commented on HBASE-9514:
--

bq.  I don't see a clear reason to wait here for every exception. So I didn't 
add it.

Ok.  I misread it as a blanket wait but now I see it just for connection issues.

Can you describe what the changes are then in a bit of text here in the issue?  
It seems like I've covered the main changes if the big bit in the middle is 
just formatting.  Thanks Jimmy. 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-16 Thread stack (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13768813#comment-13768813
 ] 

stack commented on HBASE-9514:
--

Here you are adding back the random region move to try and bring on the issue 
again:

-new FlushRandomRegionOfTableAction(tableName)
+new FlushRandomRegionOfTableAction(tableName),
+new MoveRandomRegionOfTableAction(tableName)

Why we need this?

-this.maximumAttempts =
-  
this.server.getConfiguration().getInt("hbase.assignment.maximum.attempts", 10);
+this.maximumAttempts = Math.max(1,
+  
this.server.getConfiguration().getInt("hbase.assignment.maximum.attempts", 10));

It could be configured zero?  You saying try at least once?

I suppose +  public Lock acquireLock(final String encodedName) { has to be 
public because SSH wants to use it too?

How long do servers hang out in dead servers?

{code}
+if (!region.isMetaRegion() &&
+regionStates.wasRegionOnDeadServer(encodedName)) {
+  LOG.info("Skip assigning " + region.getRegionNameAsString()
++ " because it's host " + 
regionStates.getLastRegionServerOfRegion(encodedName)
++ " is dead but not processed");
+  // Make sure the region is offline so that SSH will assign it.
+  // Need to make sure we don't race with SSH.
+  regionOffline(region);
+  return;
+}
{code}

I suppose it doesn't matter if in dead server for a long time since each server 
has a startcode?

Does this big block of new code have to go into the middle of assign?  Can it 
be broken up a little into methods that are easier to grok?

{code}
+if (serverManager.isServerOnline(server) &&
+(t instanceof java.net.SocketTimeoutException ||
+t instanceof java.net.ConnectException)) {
{code}

Is it a good idea inserting this wait here for every exception?  What if the 
exception is a NSRE?  Doesn't NSRE indicate live server?

The big change in the middle I cannot follow.  Can we have a note on what it 
does?

Do declare and assign in one go I'd say:

+lastAssignments = new HashMap();

I like this map in RS.

Good stuff Jimmy

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-16 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13768847#comment-13768847
 ] 

Jimmy Xiang commented on HBASE-9514:


bq. Here you are adding back the random region move to try and bring on the 
issue again:
Yes, I'd like to see the data loss issue in HBASE-9338 is fixed.

bq. It could be configured zero? You saying try at least once?
It can't be 0. It is total max attempts instead of retries.  I added it just in 
case users mis-configure it.

bq. I suppose + public Lock acquireLock(final String encodedName) { has to be 
public because SSH wants to use it too?
That's right.

bq. I suppose it doesn't matter if in dead server for a long time since each 
server has a startcode?
That's right. Each server has a startcode. The region will stay in 
lastAssignments till SSH log split is done.

bq. Does this big block of new code have to go into the middle of assign? Can 
it be broken up a little into methods that are easier to grok?
bq. The big change in the middle I cannot follow. Can we have a note on what it 
does?
In the middle, there is no change at all actually.  I just aligned the 
try-block.

bq. Is it a good idea inserting this wait here for every exception? What if the 
exception is a NSRE? Doesn't NSRE indicate live server?
For NSRE, the region is offlined right away. To reduce the impact, I don't see 
a clear reason to wait here for every exception. So I didn't add it.



> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
> Attachments: trunk-9514_v1.patch
>
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-16 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13768579#comment-13768579
 ] 

Enis Soztutar commented on HBASE-9514:
--

Raising this as a blocker is justified I believe. Note that both the balancer, 
and merge's are also forcing region assignment, not just user submitted region 
assignments. 
bq. So I was thinking to control it from the root, AM#assign() method, the 
final place an openRegion request is sent out to another RS.
Sounds good. I was thinking adding a boolean, isAssignable, to the master's in 
memory state. SSH will first mark the regions as unAssignable, and log 
splitting termination will reset the bit. AM#assign() will ignore regions who 
are not assignable. 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>Priority: Blocker
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-14 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13767546#comment-13767546
 ] 

Jimmy Xiang commented on HBASE-9514:


How about a RS is dead but master doesn't know about it yet? So I was thinking 
to control it from the root, AM#assign() method, the final place an openRegion 
request is sent out to another RS.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-13 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13767268#comment-13767268
 ] 

Enis Soztutar commented on HBASE-9514:
--

Thinking again, what if the balancer does those checks, then the RS is dead. We 
might still be moving the region of a dead server without split finishing from 
the balancer as far as I understand, right? 

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-13 Thread Enis Soztutar (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13767261#comment-13767261
 ] 

Enis Soztutar commented on HBASE-9514:
--

The balancer is not effected by this I think. Can we do the same checks in the 
region move code paths? 

{code:title=HMaster.balance()}
if (this.assignmentManager.getRegionStates().isRegionsInTransition()) {
Map regionsInTransition =
  this.assignmentManager.getRegionStates().getRegionsInTransition();
LOG.debug("Not running balancer because " + regionsInTransition.size() +
  " region(s) in transition: " + org.apache.commons.lang.StringUtils.
abbreviate(regionsInTransition.toString(), 256));
return false;
  }
  if (this.serverManager.areDeadServersInProgress()) {
LOG.debug("Not running balancer because processing dead 
regionserver(s): " +
  this.serverManager.getDeadServers());
return false;
  }
{code}

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-11 Thread stack (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13764946#comment-13764946
 ] 

stack commented on HBASE-9514:
--

Should be good if you include startcode.


> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

[jira] [Commented] (HBASE-9514) Prevent region from assigning before log splitting is done

2013-09-11 Thread Jimmy Xiang (JIRA)


[ 
https://issues.apache.org/jira/browse/HBASE-9514?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13764932#comment-13764932
 ] 

Jimmy Xiang commented on HBASE-9514:


I was thinking about saving the previous region server of a region in its 
region state. Whenever a region is transitioning to open, master checks if its 
previous host region server is known to be dead but not null. If so, aborting 
the open. The previous region server will be cleared only after SSH finishes 
log splitting, or master restarts of course.

> Prevent region from assigning before log splitting is done
> --
>
> Key: HBASE-9514
> URL: https://issues.apache.org/jira/browse/HBASE-9514
> Project: HBase
>  Issue Type: Bug
>  Components: Region Assignment
>Reporter: Jimmy Xiang
>Assignee: Jimmy Xiang
>
> If a region is assigned before log splitting is done by the server shutdown 
> handler, the edits belonging to this region in the hlogs of the dead server 
> will be lost.
> Generally this is not an issue if users don't assign/unassign a region from 
> hbase shell or via hbase admin. These commands are marked for experts only in 
> the hbase shell help too.  However, chaos monkey doesn't care.
> If we can prevent from assigning such regions in a bad time, it would make 
> things a little safer.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

58 matches

Mail list logo