[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Andrew Wang updated HADOOP-13826: - Fix Version/s: 3.0.0-alpha3 > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Fix For: 2.8.0, 3.0.0-alpha3 > > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch, > HADOOP-13826-branch-2-006.patch, HADOOP-13826-branch-2-007.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Resolution: Fixed Fix Version/s: 2.8.0 Status: Resolved (was: Patch Available) committed; thanks for working on this Sean > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Fix For: 2.8.0 > > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch, > HADOOP-13826-branch-2-006.patch, HADOOP-13826-branch-2-007.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Attachment: HADOOP-13826-branch-2-007.patch patch 007: checkstyle corrections to test suite. Also patched up base pathname to match renamed test classname. There's still a warning about unchecked conversion of an array; I can't make that go away (even with an attempt to suppress it), so am going to ignore that complaint > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch, > HADOOP-13826-branch-2-006.patch, HADOOP-13826-branch-2-007.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Status: Patch Available (was: Open) > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch, > HADOOP-13826-branch-2-006.patch, HADOOP-13826-branch-2-007.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Status: Open (was: Patch Available) > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch, > HADOOP-13826-branch-2-006.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Status: Patch Available (was: Open) testing, s3 ireland > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch, > HADOOP-13826-branch-2-006.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Attachment: HADOOP-13826-branch-2-006.patch Patch 006; me applying yetus feedback to the test code I'd changed. > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch, > HADOOP-13826-branch-2-006.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Status: Open (was: Patch Available) javac warning seems legit; checkstyles are all about indentation in my modified test code. Will fix. And will look at (new) IDE indentation settings > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Status: Patch Available (was: Open) > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Attachment: HADOOP-13206-branch-2-005.patch Production code LGTM (mostly); test code needed some tweaking, though the core test algorithm is good Hence: HADOOP-13286 patch 005 Production code * thread pool given a name * imports tweaked Test code: * moved test to org.apache.hadoop.fs.s3a.scale.ITestS3AConcurrentOps to emphasise scale nature & for s3a * all constant strings -> refs off Constants, to aid finding & avoid typos * timeout logic integrating with S3AScaleTestBase rules. * move off assert true/false to asserts with meaningful messages/error diagnostics * various other IDE-suggested cleanups of the test code Tested: S3a ireland @ scale. Once yetus is happy I'll put this in > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13206-branch-2-005.patch, HADOOP-13826.001.patch, > HADOOP-13826.002.patch, HADOOP-13826.003.patch, HADOOP-13826.004.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Status: Open (was: Patch Available) > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13826.001.patch, HADOOP-13826.002.patch, > HADOOP-13826.003.patch, HADOOP-13826.004.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Sean Mackrory updated HADOOP-13826: --- Attachment: HADOOP-13826.004.patch Attaching a patch that I believe addresses pretty much all your feedback. Not acting on the ideas about adding an explicit copy-only operation just yet, or the callbacks. I think that would be best addressed as a separate issue, independent of changing the resource pools? > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13826.001.patch, HADOOP-13826.002.patch, > HADOOP-13826.003.patch, HADOOP-13826.004.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Junping Du updated HADOOP-13826: Status: Patch Available (was: Open) > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13826.001.patch, HADOOP-13826.002.patch, > HADOOP-13826.003.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Sean Mackrory updated HADOOP-13826: --- Attachment: HADOOP-13826.003.patch For the sake of trying stuff out, attaching a patch that gives an unbounded ThreadPoolExecutor to the BlockingThreadPoolExecutorService, and the original unbounded one to everything else. All tests pass, including the new test that was previously able to induce a deadlock. I like [~Thomas Demoor]'s point about the control tasks not being memory intensive: having control tasks in an unbounded queue and not having to worry about them overwhelming resources too easily would solve the concern about how to make all these individual pools easily configurable. I'm fairly certain my original proposal would work more completely if rather than having 3 nested executors and only the inner-most one separating tasks into isolated pools, the outer-most executor immediately separated tasks into their own queues as well, and that would still need to be done, but there's still also the concern about relying on internal AWS APIs, which we should probably avoid. > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13826.001.patch, HADOOP-13826.002.patch, > HADOOP-13826.003.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Junping Du updated HADOOP-13826: Assignee: Sean Mackrory > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Assignee: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13826.001.patch, HADOOP-13826.002.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Target Version/s: 2.8.0 Priority: Critical (was: Major) > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory >Priority: Critical > Attachments: HADOOP-13826.001.patch, HADOOP-13826.002.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Affects Version/s: 2.7.3 > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory > Attachments: HADOOP-13826.001.patch, HADOOP-13826.002.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Summary: S3A Deadlock in multipart copy due to thread pool limits. (was: [s3a] Deadlock possible using Amazon S3 SDK) > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory > Attachments: HADOOP-13826.001.patch, HADOOP-13826.002.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
[jira] [Updated] (HADOOP-13826) S3A Deadlock in multipart copy due to thread pool limits.
[ https://issues.apache.org/jira/browse/HADOOP-13826?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steve Loughran updated HADOOP-13826: Component/s: fs/s3 > S3A Deadlock in multipart copy due to thread pool limits. > - > > Key: HADOOP-13826 > URL: https://issues.apache.org/jira/browse/HADOOP-13826 > Project: Hadoop Common > Issue Type: Bug > Components: fs/s3 >Affects Versions: 2.7.3 >Reporter: Sean Mackrory > Attachments: HADOOP-13826.001.patch, HADOOP-13826.002.patch > > > In testing HIVE-15093 we have encountered deadlocks in the s3a connector. The > TransferManager javadocs > (http://docs.aws.amazon.com/AWSJavaSDK/latest/javadoc/com/amazonaws/services/s3/transfer/TransferManager.html) > explain how this is possible: > {quote}It is not recommended to use a single threaded executor or a thread > pool with a bounded work queue as control tasks may submit subtasks that > can't complete until all sub tasks complete. Using an incorrectly configured > thread pool may cause a deadlock (I.E. the work queue is filled with control > tasks that can't finish until subtasks complete but subtasks can't execute > because the queue is filled).{quote} -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org