[jira] [Commented] (SPARK-2532) Fix issues with consolidated shuffle
[ https://issues.apache.org/jira/browse/SPARK-2532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14184576#comment-14184576 ] Patrick Wendell commented on SPARK-2532: Hey [~matei] - you created some sub-tasks here that are pretty tersely described... would you mind looking through them and deciding whether these are still relevant? Not sure whether we can close this. Fix issues with consolidated shuffle Key: SPARK-2532 URL: https://issues.apache.org/jira/browse/SPARK-2532 Project: Spark Issue Type: Bug Components: Shuffle, Spark Core Affects Versions: 1.1.0 Environment: All Reporter: Mridul Muralidharan Assignee: Mridul Muralidharan Priority: Critical Will file PR with changes as soon as merge is done (earlier merge became outdated in 2 weeks unfortunately :) ). Consolidated shuffle is broken in multiple ways in spark : a) Task failure(s) can cause the state to become inconsistent. b) Multiple revert's or combination of close/revert/close can cause the state to be inconsistent. (As part of exception/error handling). c) Some of the api in block writer causes implementation issues - for example: a revert is always followed by close : but the implemention tries to keep them separate, resulting in surface for errors. d) Fetching data from consolidated shuffle files can go badly wrong if the file is being actively written to : it computes length by subtracting next offset from current offset (or length if this is last offset)- the latter fails when fetch is happening in parallel to write. Note, this happens even if there are no task failures of any kind ! This usually results in stream corruption or decompression errors. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-2532) Fix issues with consolidated shuffle
[ https://issues.apache.org/jira/browse/SPARK-2532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14148666#comment-14148666 ] Andrew Ash commented on SPARK-2532: --- [~pwendell] should we close this ticket and track the individual items separately? It sounds like we should expect consolidated shuffle to work in 1.1 and any issues should have separate tickets filed for them. I know [~mridulm80] has several fixes on his branch that should be cherry picked over at some point as well though. Fix issues with consolidated shuffle Key: SPARK-2532 URL: https://issues.apache.org/jira/browse/SPARK-2532 Project: Spark Issue Type: Bug Components: Shuffle, Spark Core Affects Versions: 1.1.0 Environment: All Reporter: Mridul Muralidharan Assignee: Mridul Muralidharan Priority: Critical Will file PR with changes as soon as merge is done (earlier merge became outdated in 2 weeks unfortunately :) ). Consolidated shuffle is broken in multiple ways in spark : a) Task failure(s) can cause the state to become inconsistent. b) Multiple revert's or combination of close/revert/close can cause the state to be inconsistent. (As part of exception/error handling). c) Some of the api in block writer causes implementation issues - for example: a revert is always followed by close : but the implemention tries to keep them separate, resulting in surface for errors. d) Fetching data from consolidated shuffle files can go badly wrong if the file is being actively written to : it computes length by subtracting next offset from current offset (or length if this is last offset)- the latter fails when fetch is happening in parallel to write. Note, this happens even if there are no task failures of any kind ! This usually results in stream corruption or decompression errors. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-2532) Fix issues with consolidated shuffle
[ https://issues.apache.org/jira/browse/SPARK-2532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14082878#comment-14082878 ] Matei Zaharia commented on SPARK-2532: -- I'm going to create a few sub-tasks for the major improvements here to make it easier to put some of them in 1.1 and leave others for later. Fix issues with consolidated shuffle Key: SPARK-2532 URL: https://issues.apache.org/jira/browse/SPARK-2532 Project: Spark Issue Type: Bug Components: Spark Core Affects Versions: 1.1.0 Environment: All Reporter: Mridul Muralidharan Assignee: Mridul Muralidharan Priority: Critical Fix For: 1.1.0 Will file PR with changes as soon as merge is done (earlier merge became outdated in 2 weeks unfortunately :) ). Consolidated shuffle is broken in multiple ways in spark : a) Task failure(s) can cause the state to become inconsistent. b) Multiple revert's or combination of close/revert/close can cause the state to be inconsistent. (As part of exception/error handling). c) Some of the api in block writer causes implementation issues - for example: a revert is always followed by close : but the implemention tries to keep them separate, resulting in surface for errors. d) Fetching data from consolidated shuffle files can go badly wrong if the file is being actively written to : it computes length by subtracting next offset from current offset (or length if this is last offset)- the latter fails when fetch is happening in parallel to write. Note, this happens even if there are no task failures of any kind ! This usually results in stream corruption or decompression errors. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (SPARK-2532) Fix issues with consolidated shuffle
[ https://issues.apache.org/jira/browse/SPARK-2532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14080391#comment-14080391 ] Apache Spark commented on SPARK-2532: - User 'aarondav' has created a pull request for this issue: https://github.com/apache/spark/pull/1678 Fix issues with consolidated shuffle Key: SPARK-2532 URL: https://issues.apache.org/jira/browse/SPARK-2532 Project: Spark Issue Type: Bug Components: Spark Core Affects Versions: 1.1.0 Environment: All Reporter: Mridul Muralidharan Assignee: Mridul Muralidharan Priority: Critical Fix For: 1.1.0 Will file PR with changes as soon as merge is done (earlier merge became outdated in 2 weeks unfortunately :) ). Consolidated shuffle is broken in multiple ways in spark : a) Task failure(s) can cause the state to become inconsistent. b) Multiple revert's or combination of close/revert/close can cause the state to be inconsistent. (As part of exception/error handling). c) Some of the api in block writer causes implementation issues - for example: a revert is always followed by close : but the implemention tries to keep them separate, resulting in surface for errors. d) Fetching data from consolidated shuffle files can go badly wrong if the file is being actively written to : it computes length by subtracting next offset from current offset (or length if this is last offset)- the latter fails when fetch is happening in parallel to write. Note, this happens even if there are no task failures of any kind ! This usually results in stream corruption or decompression errors. -- This message was sent by Atlassian JIRA (v6.2#6252)
[jira] [Commented] (SPARK-2532) Fix issues with consolidated shuffle
[ https://issues.apache.org/jira/browse/SPARK-2532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14075639#comment-14075639 ] Apache Spark commented on SPARK-2532: - User 'mridulm' has created a pull request for this issue: https://github.com/apache/spark/pull/1609 Fix issues with consolidated shuffle Key: SPARK-2532 URL: https://issues.apache.org/jira/browse/SPARK-2532 Project: Spark Issue Type: Bug Components: Spark Core Affects Versions: 1.1.0 Environment: All Reporter: Mridul Muralidharan Assignee: Mridul Muralidharan Priority: Critical Fix For: 1.1.0 Will file PR with changes as soon as merge is done (earlier merge became outdated in 2 weeks unfortunately :) ). Consolidated shuffle is broken in multiple ways in spark : a) Task failure(s) can cause the state to become inconsistent. b) Multiple revert's or combination of close/revert/close can cause the state to be inconsistent. (As part of exception/error handling). c) Some of the api in block writer causes implementation issues - for example: a revert is always followed by close : but the implemention tries to keep them separate, resulting in surface for errors. d) Fetching data from consolidated shuffle files can go badly wrong if the file is being actively written to : it computes length by subtracting next offset from current offset (or length if this is last offset)- the latter fails when fetch is happening in parallel to write. Note, this happens even if there are no task failures of any kind ! This usually results in stream corruption or decompression errors. -- This message was sent by Atlassian JIRA (v6.2#6252)