[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13411480#comment-13411480 ] Hudson commented on MAPREDUCE-3728: --- Integrated in Hadoop-Hdfs-0.23-Build #310 (See [https://builds.apache.org/job/Hadoop-Hdfs-0.23-Build/310/]) svn merge -c 1295245 FIXES: MAPREDUCE-3728. ShuffleHandler can't access results when configured in a secure mode (ahmed via tucu) (Revision 1359724) Result = UNSTABLE bobby : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1359724 Files : * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/ContainerLocalizer.java * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/test/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/TestContainerLocalizer.java > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.3, 2.0.0-alpha > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13220041#comment-13220041 ] Hudson commented on MAPREDUCE-3728: --- Integrated in Hadoop-Mapreduce-trunk #1006 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk/1006/]) MAPREDUCE-3728. ShuffleHandler can't access results when configured in a secure mode (ahmed via tucu) (Revision 1295245) Result = SUCCESS tucu : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1295245 Files : * /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/ContainerLocalizer.java * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/test/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/TestContainerLocalizer.java > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.3 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219993#comment-13219993 ] Hudson commented on MAPREDUCE-3728: --- Integrated in Hadoop-Hdfs-trunk #971 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk/971/]) MAPREDUCE-3728. ShuffleHandler can't access results when configured in a secure mode (ahmed via tucu) (Revision 1295245) Result = SUCCESS tucu : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1295245 Files : * /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/ContainerLocalizer.java * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/test/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/TestContainerLocalizer.java > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.3 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219561#comment-13219561 ] Hudson commented on MAPREDUCE-3728: --- Integrated in Hadoop-Mapreduce-trunk-Commit #1818 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk-Commit/1818/]) MAPREDUCE-3728. ShuffleHandler can't access results when configured in a secure mode (ahmed via tucu) (Revision 1295245) Result = ABORTED tucu : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1295245 Files : * /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/ContainerLocalizer.java * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/test/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/TestContainerLocalizer.java > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.3 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219557#comment-13219557 ] Hudson commented on MAPREDUCE-3728: --- Integrated in Hadoop-Mapreduce-0.23-Commit #620 (See [https://builds.apache.org/job/Hadoop-Mapreduce-0.23-Commit/620/]) Merge -r 1295244:1295245 from trunk to branch. FIXES: MAPREDUCE-3728 (Revision 1295246) Result = ABORTED tucu : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1295246 Files : * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/ContainerLocalizer.java * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/test/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/TestContainerLocalizer.java > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.3 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219499#comment-13219499 ] Hudson commented on MAPREDUCE-3728: --- Integrated in Hadoop-Hdfs-0.23-Commit #609 (See [https://builds.apache.org/job/Hadoop-Hdfs-0.23-Commit/609/]) Merge -r 1295244:1295245 from trunk to branch. FIXES: MAPREDUCE-3728 (Revision 1295246) Result = SUCCESS tucu : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1295246 Files : * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/ContainerLocalizer.java * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/test/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/TestContainerLocalizer.java > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.3 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219498#comment-13219498 ] Hudson commented on MAPREDUCE-3728: --- Integrated in Hadoop-Hdfs-trunk-Commit #1886 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk-Commit/1886/]) MAPREDUCE-3728. ShuffleHandler can't access results when configured in a secure mode (ahmed via tucu) (Revision 1295245) Result = SUCCESS tucu : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1295245 Files : * /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/ContainerLocalizer.java * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/test/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/TestContainerLocalizer.java > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.3 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219493#comment-13219493 ] Hudson commented on MAPREDUCE-3728: --- Integrated in Hadoop-Common-0.23-Commit #620 (See [https://builds.apache.org/job/Hadoop-Common-0.23-Commit/620/]) Merge -r 1295244:1295245 from trunk to branch. FIXES: MAPREDUCE-3728 (Revision 1295246) Result = SUCCESS tucu : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1295246 Files : * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/ContainerLocalizer.java * /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/test/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/TestContainerLocalizer.java > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.3 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219487#comment-13219487 ] Hudson commented on MAPREDUCE-3728: --- Integrated in Hadoop-Common-trunk-Commit #1811 (See [https://builds.apache.org/job/Hadoop-Common-trunk-Commit/1811/]) MAPREDUCE-3728. ShuffleHandler can't access results when configured in a secure mode (ahmed via tucu) (Revision 1295245) Result = SUCCESS tucu : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1295245 Files : * /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/ContainerLocalizer.java * /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/test/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/localizer/TestContainerLocalizer.java > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.3 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219474#comment-13219474 ] Alejandro Abdelnur commented on MAPREDUCE-3728: --- +1. I cannot think of any side effect by this change. Please open a JIRA against the Local impl of the LoginContext regarding the lost of the gid. > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.1 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219461#comment-13219461 ] Ahmed Radwan commented on MAPREDUCE-3728: - As I mentioned earlier, it seems that the problem is mainly on how the output directory is created in ContainerLocalizer.java. The setgid is lost in the process. I have even tried to set the permissions explicitly and setting setgid but also this doesn't seem to solve the issue. So I think the current patch, where the output directory creation is postponed is appropriate as it is correctly created afterwards by FileSystem.create(). We may also need to investigate more why mkdir in AbstractFileSystem in this case is causing the loss of setgid, and see if this a bug or it is an intentional behavior. > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.1 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13218324#comment-13218324 ] Roman Shaposhnik commented on MAPREDUCE-3728: - This patch seems to be working in my case. I'd recommend including it in the trunk and .23 branch. > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.1 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13214553#comment-13214553 ] Hadoop QA commented on MAPREDUCE-3728: -- +1 overall. Here are the results of testing the latest attachment http://issues.apache.org/jira/secure/attachment/12515738/MAPREDUCE-3728.patch against trunk revision . +1 @author. The patch does not contain any @author tags. +1 tests included. The patch appears to include 3 new or modified tests. +1 javadoc. The javadoc tool did not generate any warning messages. +1 javac. The applied patch does not increase the total number of javac compiler warnings. +1 eclipse:eclipse. The patch built with eclipse:eclipse. +1 findbugs. The patch does not introduce any new Findbugs (version 1.3.9) warnings. +1 release audit. The applied patch does not increase the total number of release audit warnings. +1 core tests. The patch passed unit tests in . +1 contrib tests. The patch passed contrib unit tests. Test results: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/1914//testReport/ Console output: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/1914//console This message is automatically generated. > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik >Assignee: Ahmed Radwan >Priority: Critical > Fix For: 0.23.1 > > Attachments: MAPREDUCE-3728.patch > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13204799#comment-13204799 ] Roman Shaposhnik commented on MAPREDUCE-3728: - Here's a more direct way to reproduce the problem. {noformat} # sudo - yarn yarn$ mkdir -p /tmp/TEST/{logs,locs} /tmp/TEST/locs/usercache yarn$ cp /tmp/cont1.tokens /tmp/TEST/cont1.tokens yarn$ container-executor rvs 0 app1 /tmp/TEST/cont1.tokens /tmp/TEST/locs /tmp/TEST/logs /usr/java/jdk1.6.0_26/jre/bin/java -classpath /usr/lib/hadoop/lib/\*:/usr/lib/hadoop/\*:/etc/hadoop/conf/nm-config/log4j.properties:/etc/hadoop/conf -Djava.library.path=/usr/lib/hadoop/lib/native org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer rvs app1 cont1 0.0.0.0 4344 /tmp/TEST/locs main : command provided 0 main : user is rvs 12/02/09 11:54:40 WARN conf.Configuration: mapred-site.xml:an attempt to override final parameter: mapreduce.cluster.local.dir; Ignoring. 12/02/09 11:54:40 WARN conf.Configuration: mapred-site.xml:an attempt to override final parameter: mapreduce.cluster.local.dir; Ignoring. 12/02/09 11:54:40 WARN conf.Configuration: mapred-site.xml:an attempt to override final parameter: mapreduce.cluster.local.dir; Ignoring. 12/02/09 11:54:40 INFO util.NativeCodeLoader: Loaded the native-hadoop library 12/02/09 11:54:40 WARN conf.Configuration: mapred-site.xml:an attempt to override final parameter: mapreduce.cluster.local.dir; Ignoring. 12/02/09 11:54:40 WARN conf.Configuration: mapred-site.xml:an attempt to override final parameter: mapreduce.cluster.local.dir; Ignoring. 12/02/09 11:54:40 WARN conf.Configuration: mapred-site.xml:an attempt to override final parameter: mapreduce.cluster.local.dir; Ignoring. === Using localizerTokenSecurityInfo12/02/09 11:54:41 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 0 time(s). 12/02/09 11:54:42 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 1 time(s). 12/02/09 11:54:43 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 2 time(s). 12/02/09 11:54:44 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 3 time(s). 12/02/09 11:54:45 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 4 time(s). 12/02/09 11:54:46 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 5 time(s). 12/02/09 11:54:47 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 6 time(s). 12/02/09 11:54:48 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 7 time(s). 12/02/09 11:54:49 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 8 time(s). 12/02/09 11:54:50 INFO ipc.Client: Retrying connect to server: 0.0.0.0/0.0.0.0:4344. Already tried 9 time(s). java.lang.reflect.UndeclaredThrowableException at org.apache.hadoop.yarn.server.nodemanager.api.impl.pb.client.LocalizationProtocolPBClientImpl.heartbeat(LocalizationProtocolPBClientImpl.java:62) at org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.localizeFiles(ContainerLocalizer.java:221) at org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.runLocalization(ContainerLocalizer.java:169) at org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.main(ContainerLocalizer.java:345) Caused by: com.google.protobuf.ServiceException: java.net.ConnectException: Call From c0506.hal.cloudera.com/172.29.81.158 to 0.0.0.0:4344 failed on connection exception: java.net.ConnectException: Connection refused; For more details see: http://wiki.apache.org/hadoop/ConnectionRefused at org.apache.hadoop.yarn.ipc.ProtoOverHadoopRpcEngine$Invoker.invoke(ProtoOverHadoopRpcEngine.java:148) at $Proxy6.heartbeat(Unknown Source) at org.apache.hadoop.yarn.server.nodemanager.api.impl.pb.client.LocalizationProtocolPBClientImpl.heartbeat(LocalizationProtocolPBClientImpl.java:54) ... 3 more Caused by: java.net.ConnectException: Call From c0506.hal.cloudera.com/172.29.81.158 to 0.0.0.0:4344 failed on connection exception: java.net.ConnectException: Connection refused; For more details see: http://wiki.apache.org/hadoop/ConnectionRefused at org.apache.hadoop.net.NetUtils.wrapException(NetUtils.java:686) at org.apache.hadoop.ipc.Client.call(Client.java:1141) at org.apache.hadoop.ipc.Client.call(Client.java:1100) at org.apache.hadoop.yarn.ipc.ProtoOverHadoopRpcEngine$Invoker.invoke(ProtoOverHadoopRpcEngine.java:145) ... 5 more Caused by: java.net.ConnectException: Connection refused at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method) at sun.nio.ch.Socket
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13193377#comment-13193377 ] Roman Shaposhnik commented on MAPREDUCE-3728: - .bq May be I am missing something, but I don't understand this. The output-dir is being created with null FilePermissions which becomes 0777. I suspect this is not about *what* permissions it is create with, but the sheer fact that permissions are set at all. Here's an example of how it goes wrong when permissions are set explicitly. When that happens sticky bit gets lost. Here's a trivial example. As you can see, when a subdir is created it is, at first, retains the sticky bit. But when I set permissions explicitly it gets cleared (as it should): {noformat} rvs@ahmed-laptop:/tmp$ cd /tmp rvs@ahmed-laptop:/tmp$ mkdir sticky.dir rvs@ahmed-laptop:/tmp$ sudo chgrp root sticky.dir rvs@ahmed-laptop:/tmp$ sudo chmod g+s sticky.dir rvs@ahmed-laptop:/tmp$ mkdir sticky.dir/subdir rvs@ahmed-laptop:/tmp$ ls -l sticky.dir total 4 drwxr-sr-x 2 rvs root 4096 2012-01-25 13:50 subdir rvs@ahmed-laptop:/tmp$ chmod 777 sticky.dir/subdir rvs@ahmed-laptop:/tmp$ ls -l sticky.dir total 4 drwxrwxrwx 2 rvs root 4096 2012-01-25 13:50 subdir {noformat} .bq Instead I suspect that the umask for your testuser is strict. That my be the reason why the files are getting created with 640. Can you please check? The problem is not file permissions per se, but the fact that subdirs (starting from output) have lost a group sticky bit and thus everything that is created underneath them has user group as a default (and of course no longer retains the sticky group bit on the subdirectories). But anyway, I've checked and the umask for testuser is 0022 .bq Also, just to be sure, you are using DefaultContainerExecutor or LinuxContainerExecutor? It is mostly the later, but just confirming. This is LinuxContainerExecutor since, to the best of my knowledge, one has to configure it for the fully secure cluster. > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik > Fix For: 0.23.1 > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited
[jira] [Commented] (MAPREDUCE-3728) ShuffleHandler can't access results when configured in a secure mode
[ https://issues.apache.org/jira/browse/MAPREDUCE-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13193358#comment-13193358 ] Vinod Kumar Vavilapalli commented on MAPREDUCE-3728: bq. Reveals that lfs.mkdir ends up manipulating permissions and thus clears sticky bit from output and filecache. May be I am missing something, but I don't understand this. The output-dir is being created with null FilePermissions which becomes 0777. Instead I suspect that the umask for your testuser is strict. That my be the reason why the files are getting created with 640. Can you please check? Also, just to be sure, you are using DefaultContainerExecutor or LinuxContainerExecutor? It is mostly the later, but just confirming. bq. My understanding was that the whole sequence of events here was predicated on a sticky bit set so that daemons running under the user yarn (default group yarn) can have access to the resulting files and subdirectories down at output and below. Yes, that is the case with YARN and even mrv1/1.0. bq. On a related note, when the shuffle side of the Pi job failed the job itself didn't. It went into the endless loop [..] Yes, this is a known issue MAPREDUCE-3418. > ShuffleHandler can't access results when configured in a secure mode > > > Key: MAPREDUCE-3728 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3728 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager >Affects Versions: 0.23.0 >Reporter: Roman Shaposhnik > Fix For: 0.23.1 > > > While running the simplest of jobs (Pi) on MR2 in a fully secure > configuration I have noticed that the job was failing on the reduce side with > the following messages littering the nodemanager logs: > {noformat} > 2012-01-19 08:35:32,544 ERROR org.apache.hadoop.mapred.ShuffleHandler: > Shuffle error > org.apache.hadoop.util.DiskChecker$DiskErrorException: Could not find > usercache/rvs/appcache/application_1326928483038_0001/output/attempt_1326928483038_0001_m_03_0/file.out.index > in any of the configured local directories > {noformat} > While digging further I found out that the permissions on the files/dirs were > prohibiting nodemanager (running under the user yarn) to access these files: > {noformat} > $ ls -l > /data/3/yarn/usercache/testuser/appcache/application_1327102703969_0001/output/attempt_1327102703969_0001_m_01_0 > -rw-r- 1 testuser testuser 28 Jan 20 15:41 file.out > -rw-r- 1 testuser testuser 32 Jan 20 15:41 file.out.index > {noformat} > Digging even further revealed that the group-sticky bit that was faithfully > put on all the subdirectories between testuser and > application_1327102703969_0001 was gone from output and > attempt_1327102703969_0001_m_01_0. > Looking into how these subdirectories are created > (org.apache.hadoop.yarn.server.nodemanager.containermanager.localizer.ContainerLocalizer.initDirs()) > {noformat} > // $x/usercache/$user/appcache/$appId/filecache > Path appFileCacheDir = new Path(appBase, FILECACHE); > appsFileCacheDirs[i] = appFileCacheDir.toString(); > lfs.mkdir(appFileCacheDir, null, false); > // $x/usercache/$user/appcache/$appId/output > lfs.mkdir(new Path(appBase, OUTPUTDIR), null, false); > {noformat} > Reveals that lfs.mkdir ends up manipulating permissions and thus clears > sticky bit from output and filecache. > At this point I'm at a loss about how this is supposed to work. My > understanding was > that the whole sequence of events here was predicated on a sticky bit set so > that daemons running under the user yarn (default group yarn) can have access > to the resulting files and subdirectories down at output and below. Please let > me know if I'm missing something or whether this is just a bug that needs to > be fixed. > On a related note, when the shuffle side of the Pi job failed the job itself > didn't. > It went into the endless loop and only exited when it exhausted all the local > storage > for the log files (at which point the nodemanager died and thus the job > ended). Perhaps > this is even more serious side effect of this issue that needs to be > investigated > separately. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira