[jira] [Commented] (YARN-9593) Updating scheduler conf with comma in config value fails
[ https://issues.apache.org/jira/browse/YARN-9593?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16860396#comment-16860396 ] Anthony Hsu commented on YARN-9593: --- Thanks, [~jhung]. I don't have bandwidth at the moment, but glad we're agreed on the approach. I think this would be a good starter task. > Updating scheduler conf with comma in config value fails > > > Key: YARN-9593 > URL: https://issues.apache.org/jira/browse/YARN-9593 > Project: Hadoop YARN > Issue Type: Bug >Affects Versions: 2.9.0, 3.0.0, 3.2.0, 3.1.2 >Reporter: Anthony Hsu >Priority: Major > > For example: > {code:java} > $ yarn schedulerconf -update "root.gridops:acl_administer_queue=user1,user2 > group1,group2" > Specify configuration key value as confKey=confVal.{code} > This fails because there is a comma in the config value and the SchedConfCLI > splits on comma first, expecting each split to a k=v pair. > {noformat} > void globalUpdates(String args, SchedConfUpdateInfo updateInfo) { > if (args == null) { > return; > } > HashMap globalUpdates = new HashMap<>(); > for (String globalUpdate : args.split(",")) { > putKeyValuePair(globalUpdates, globalUpdate); > } > updateInfo.setGlobalParams(globalUpdates); > }{noformat} > Cc: [~jhung] -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (YARN-3213) Respect labels in Capacity Scheduler when computing user-limit
[ https://issues.apache.org/jira/browse/YARN-3213?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16857180#comment-16857180 ] Anthony Hsu edited comment on YARN-3213 at 6/6/19 12:37 AM: Thanks for confirming, Wangda. Created YARN-9603 for updating CS to support specifying different minimum-user-limit-percent and user-limit-factor on a per node label basis. was (Author: erwaman): Thanks for confirming, Wangda. Will file a Jira for updating CS to support specifying different minimum-user-limit-percent and user-limit-factor on a per node label basis. > Respect labels in Capacity Scheduler when computing user-limit > -- > > Key: YARN-3213 > URL: https://issues.apache.org/jira/browse/YARN-3213 > Project: Hadoop YARN > Issue Type: Sub-task > Components: capacityscheduler >Reporter: Wangda Tan >Assignee: Wangda Tan >Priority: Major > > Now we can support node-labels in Capacity Scheduler, but user-limit > computing doesn't respect node-labels enough, we should fix that. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Created] (YARN-9603) Support setting user limit percent and user limit factor for a queue on a per node label basis
Anthony Hsu created YARN-9603: - Summary: Support setting user limit percent and user limit factor for a queue on a per node label basis Key: YARN-9603 URL: https://issues.apache.org/jira/browse/YARN-9603 Project: Hadoop YARN Issue Type: New Feature Reporter: Anthony Hsu User limit settings (minimum-user-limit-percent, user-limit-factor) for queues can NOT be set on a per node label basis. However, capacity settings (capacity, maximum-capacity) CAN be set on a per node label basis. This Jira is for supporting setting different user limits per queue per node label. For more context, see discussion in YARN-3213. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-3213) Respect labels in Capacity Scheduler when computing user-limit
[ https://issues.apache.org/jira/browse/YARN-3213?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16857180#comment-16857180 ] Anthony Hsu commented on YARN-3213: --- Thanks for confirming, Wangda. Will file a Jira for updating CS to support specifying different minimum-user-limit-percent and user-limit-factor on a per node label basis. > Respect labels in Capacity Scheduler when computing user-limit > -- > > Key: YARN-3213 > URL: https://issues.apache.org/jira/browse/YARN-3213 > Project: Hadoop YARN > Issue Type: Sub-task > Components: capacityscheduler >Reporter: Wangda Tan >Assignee: Wangda Tan >Priority: Major > > Now we can support node-labels in Capacity Scheduler, but user-limit > computing doesn't respect node-labels enough, we should fix that. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-3213) Respect labels in Capacity Scheduler when computing user-limit
[ https://issues.apache.org/jira/browse/YARN-3213?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16857020#comment-16857020 ] Anthony Hsu commented on YARN-3213: --- Hi [~leftnoteasy], It seems even after YARN-3361, you are unable to set user-limit for a queue on a per-label basis. In CapacitySchedulerConfiguration, I see * getLabeledQueueCapacity / setCapacityByLabel * getLabeledQueueMaximumCapacity / setMaximumCapacityByLabel but I don't see corresponding per-label methods for user limit and user limit factor. Is this intentional? > Respect labels in Capacity Scheduler when computing user-limit > -- > > Key: YARN-3213 > URL: https://issues.apache.org/jira/browse/YARN-3213 > Project: Hadoop YARN > Issue Type: Sub-task > Components: capacityscheduler >Reporter: Wangda Tan >Assignee: Wangda Tan >Priority: Major > > Now we can support node-labels in Capacity Scheduler, but user-limit > computing doesn't respect node-labels enough, we should fix that. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9593) Updating scheduler conf with comma in config value fails
[ https://issues.apache.org/jira/browse/YARN-9593?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16853546#comment-16853546 ] Anthony Hsu commented on YARN-9593: --- Hi [~jhung], Thanks for sharing the workaround. That works for me for now. I think the CLI can be updated to ignore escaped commas ({{\,}}) to support confs with commas (which can escape the commas). How does that sound to you? > Updating scheduler conf with comma in config value fails > > > Key: YARN-9593 > URL: https://issues.apache.org/jira/browse/YARN-9593 > Project: Hadoop YARN > Issue Type: Bug >Affects Versions: 2.9.0, 3.0.0, 3.2.0, 3.1.2 >Reporter: Anthony Hsu >Priority: Major > > For example: > {code:java} > $ yarn schedulerconf -update "root.gridops:acl_administer_queue=user1,user2 > group1,group2" > Specify configuration key value as confKey=confVal.{code} > This fails because there is a comma in the config value and the SchedConfCLI > splits on comma first, expecting each split to a k=v pair. > {noformat} > void globalUpdates(String args, SchedConfUpdateInfo updateInfo) { > if (args == null) { > return; > } > HashMap globalUpdates = new HashMap<>(); > for (String globalUpdate : args.split(",")) { > putKeyValuePair(globalUpdates, globalUpdate); > } > updateInfo.setGlobalParams(globalUpdates); > }{noformat} > Cc: [~jhung] -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Created] (YARN-9593) Updating scheduler conf with comma in config value fails
Anthony Hsu created YARN-9593: - Summary: Updating scheduler conf with comma in config value fails Key: YARN-9593 URL: https://issues.apache.org/jira/browse/YARN-9593 Project: Hadoop YARN Issue Type: Bug Affects Versions: 3.1.2, 3.2.0, 3.0.0, 2.9.0 Reporter: Anthony Hsu For example: {code:java} $ yarn schedulerconf -update "root.gridops:acl_administer_queue=user1,user2 group1,group2" Specify configuration key value as confKey=confVal.{code} This fails because there is a comma in the config value and the SchedConfCLI splits on comma first, expecting each split to a k=v pair. {noformat} void globalUpdates(String args, SchedConfUpdateInfo updateInfo) { if (args == null) { return; } HashMap globalUpdates = new HashMap<>(); for (String globalUpdate : args.split(",")) { putKeyValuePair(globalUpdates, globalUpdate); } updateInfo.setGlobalParams(globalUpdates); }{noformat} Cc: [~jhung] -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9559) Create AbstractContainersLauncher for pluggable ContainersLauncher logic
[ https://issues.apache.org/jira/browse/YARN-9559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16841494#comment-16841494 ] Anthony Hsu commented on YARN-9559: --- +1 on your patch (non-binding). > Create AbstractContainersLauncher for pluggable ContainersLauncher logic > > > Key: YARN-9559 > URL: https://issues.apache.org/jira/browse/YARN-9559 > Project: Hadoop YARN > Issue Type: Task >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9559.001.patch > > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9529) Log correct cpu controller path on error
[ https://issues.apache.org/jira/browse/YARN-9529?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16832958#comment-16832958 ] Anthony Hsu commented on YARN-9529: --- +1 (non-binding). What were you trying to debug that requires this change? > Log correct cpu controller path on error > > > Key: YARN-9529 > URL: https://issues.apache.org/jira/browse/YARN-9529 > Project: Hadoop YARN > Issue Type: Improvement >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9529.001.patch > > > The base cpu controller path is logged instead of the hadoop cgroup path. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9412) Backport YARN-6909 to branch-2
[ https://issues.apache.org/jira/browse/YARN-9412?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16803465#comment-16803465 ] Anthony Hsu commented on YARN-9412: --- Awesome! > Backport YARN-6909 to branch-2 > -- > > Key: YARN-9412 > URL: https://issues.apache.org/jira/browse/YARN-9412 > Project: Hadoop YARN > Issue Type: Sub-task >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (YARN-9409) Port resource type changes from YARN-7237 to branch-3.0/branch-2
[ https://issues.apache.org/jira/browse/YARN-9409?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16802136#comment-16802136 ] Anthony Hsu edited comment on YARN-9409 at 3/26/19 8:27 PM: +1 (non-binding). Can you also backport YARN-6909 which is needed to write end-to-end unit tests that involve request new resource types like GPUs? It probably makes sense to do this in a separate subtask of YARN-8200. I'd be happy to review the patch. was (Author: erwaman): +1 (non-binding). Can you also backport YARN-6909 which is needed to write end-to-end unit tests that involve request new resource types like GPUs? > Port resource type changes from YARN-7237 to branch-3.0/branch-2 > > > Key: YARN-9409 > URL: https://issues.apache.org/jira/browse/YARN-9409 > Project: Hadoop YARN > Issue Type: Sub-task >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9409-YARN-8200.branch3.001.patch > > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (YARN-9409) Port resource type changes from YARN-7237 to branch-3.0/branch-2
[ https://issues.apache.org/jira/browse/YARN-9409?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16802136#comment-16802136 ] Anthony Hsu edited comment on YARN-9409 at 3/26/19 8:27 PM: +1 (non-binding). Can you also backport YARN-6909 which is needed to write end-to-end unit tests that involve request new resource types like GPUs? It probably makes sense to do this in a separate subtask of YARN-8200. I'd be happy to review the patch. was (Author: erwaman): +1 (non-binding). Can you also backport YARN-6909 which is needed to write end-to-end unit tests that involve request new resource types like GPUs? It probably makes sense to do this in a separate subtask of YARN-8200. I'd be happy to review the patch. > Port resource type changes from YARN-7237 to branch-3.0/branch-2 > > > Key: YARN-9409 > URL: https://issues.apache.org/jira/browse/YARN-9409 > Project: Hadoop YARN > Issue Type: Sub-task >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9409-YARN-8200.branch3.001.patch > > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9409) Port resource type changes from YARN-7237 to branch-3.0/branch-2
[ https://issues.apache.org/jira/browse/YARN-9409?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16802136#comment-16802136 ] Anthony Hsu commented on YARN-9409: --- +1 (non-binding). Can you also backport YARN-6909 which is needed to write end-to-end unit tests that involve request new resource types like GPUs? > Port resource type changes from YARN-7237 to branch-3.0/branch-2 > > > Key: YARN-9409 > URL: https://issues.apache.org/jira/browse/YARN-9409 > Project: Hadoop YARN > Issue Type: Sub-task >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9409-YARN-8200.branch3.001.patch > > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9272) Backport YARN-7738 for refreshing max allocation for multiple resource types
[ https://issues.apache.org/jira/browse/YARN-9272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16802133#comment-16802133 ] Anthony Hsu commented on YARN-9272: --- I looked at the branch-2 patch. It looks good. +1 (non-binding). > Backport YARN-7738 for refreshing max allocation for multiple resource types > > > Key: YARN-9272 > URL: https://issues.apache.org/jira/browse/YARN-9272 > Project: Hadoop YARN > Issue Type: Sub-task >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9272-YARN-8200.001.patch, > YARN-9272-YARN-8200.branch3.001.patch, YARN-9272-YARN-8200.branch3.002.patch > > > Need to port to YARN-8200.branch3 (for branch-3.0) and YARN-8200 (for > branch-2) -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-6686) Support for adding and removing queue mappings
[ https://issues.apache.org/jira/browse/YARN-6686?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16768763#comment-16768763 ] Anthony Hsu commented on YARN-6686: --- Hi [~csexz], I like your proposal of using and . I have one suggestion. Instead of having separate and , I think everything should be an ( to me sounds like it's a read-only request, not an update request). * If just contains text, then it's a complete replacement. * If has and children, then it's an addition, removal, or update ||Operation|| Description/Notes|| |Addition|(empty)|u::,u::,...|Adds the mappings u::,u::,... | |Removal|u::,u::,...|(empty)|Removes the mappings u::,u::,...| |Update|u::,u::,...|u::,u::,...|Updates the mappings for users , , ... from , , ... to , , ... The and should contain the same number of items.| > Support for adding and removing queue mappings > -- > > Key: YARN-6686 > URL: https://issues.apache.org/jira/browse/YARN-6686 > Project: Hadoop YARN > Issue Type: Sub-task >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-6686-YARN-5734.001.patch > > > Right now capacity scheduler uses UserGroupMappingPlacementRule to determine > queue mappings. This rule stores mappings in > {{yarn.scheduler.capacity.queue-mappings}}. For users with a large number of > mappings, adding or removing queue mappings becomes infeasible. > Need to come up with a way to add/remove individual mappings, for any/all > different configured placement rules. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9174) branch-3.0/branch-2 refactoring of GpuDevice class
[ https://issues.apache.org/jira/browse/YARN-9174?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16736133#comment-16736133 ] Anthony Hsu commented on YARN-9174: --- What is the motivation for backporting the GpuDevice class refactoring to branch-2? Is there another Jira that depends on this? > branch-3.0/branch-2 refactoring of GpuDevice class > -- > > Key: YARN-9174 > URL: https://issues.apache.org/jira/browse/YARN-9174 > Project: Hadoop YARN > Issue Type: Sub-task >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9174-YARN-8200.001.patch, > YARN-9174-YARN-8200.branch3.001.patch, YARN-9174-YARN-8200.branch3.002.patch > > > YARN-7224 does two main things: > # refactors Gpu device numbers to a separate GpuDevice class, > # adds Docker support for Gpus > This ticket is for doing *only* the GpuDevice class refactoring so we have > this logic in branch-3.0 and branch-2 -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9085) Guaranteed and MaxCapacity CSQueueMetrics
[ https://issues.apache.org/jira/browse/YARN-9085?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16713098#comment-16713098 ] Anthony Hsu commented on YARN-9085: --- Latest patch looks good to me. +1 (non-binding) > Guaranteed and MaxCapacity CSQueueMetrics > - > > Key: YARN-9085 > URL: https://issues.apache.org/jira/browse/YARN-9085 > Project: Hadoop YARN > Issue Type: Improvement >Affects Versions: 2.9.3 >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9085.001.patch, YARN-9085.002.patch > > > Would be useful to have Absolute Capacity/Absolute Max Capacity for queues to > compare against allocated/pending/etc. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9085) Guaranteed and MaxCapacity CSQueueMetrics
[ https://issues.apache.org/jira/browse/YARN-9085?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16713094#comment-16713094 ] Anthony Hsu commented on YARN-9085: --- {quote}GPU metrics should be addressable in YARN-8842 (but this is currently only in 3.3.0) {quote} Can we backport this to branch 2.x? GPU metrics are important for those doing deep learning on GPUs. > Guaranteed and MaxCapacity CSQueueMetrics > - > > Key: YARN-9085 > URL: https://issues.apache.org/jira/browse/YARN-9085 > Project: Hadoop YARN > Issue Type: Improvement >Affects Versions: 2.9.3 >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9085.001.patch, YARN-9085.002.patch > > > Would be useful to have Absolute Capacity/Absolute Max Capacity for queues to > compare against allocated/pending/etc. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-9085) Guaranteed and MaxCapacity CSQueueMetrics
[ https://issues.apache.org/jira/browse/YARN-9085?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711709#comment-16711709 ] Anthony Hsu commented on YARN-9085: --- Looks good overall. A few questions/comments: * In CSQueueMetrics.java, I see the comment {code:java} //Metrics updated only for "default" partition{code} How come metrics are not updated for non-default partitions? Are any metrics available for non-default partitions? * What about GPU metrics? Can CSQueueMetrics collect those, too? * Regarding {noformat} if (nodePartition == null) { for (String partition : Sets.union(queueCapacities.getNodePartitionsSet(), queueResourceUsage.getNodePartitionsSet())) { -updateUsedCapacity(rc, nlm.getResourceByLabel(partition, cluster), -partition, childQueue); +updateUsedCapacity(rc, partitionResource, partition, childQueue); } + updateConfiguredCapacities(rc, partitionResource, childQueue); } else { - updateUsedCapacity(rc, nlm.getResourceByLabel(nodePartition, cluster), - nodePartition, childQueue); + updateUsedCapacity(rc, partitionResource, nodePartition, childQueue); } {noformat} Seems to me the *updateConfiguredCapacities* call you added should be inside the for loop and should also take in a *partition* parameter like the *updateUsedCapacity* call does. In the future, metrics may be collected for non-default partitions as well. Also, I think the *else* block should also have a *updateConfiguredCapacities* call (in case in the future, we collect non-default partition metrics, too. > Guaranteed and MaxCapacity CSQueueMetrics > - > > Key: YARN-9085 > URL: https://issues.apache.org/jira/browse/YARN-9085 > Project: Hadoop YARN > Issue Type: Improvement >Affects Versions: 2.9.3 >Reporter: Jonathan Hung >Assignee: Jonathan Hung >Priority: Major > Attachments: YARN-9085.001.patch > > > Would be useful to have Absolute Capacity/Absolute Max Capacity for queues to > compare against allocated/pending/etc. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org