[jira] [Commented] (AURORA-1780) Offers with unknown resources types to Aurora crash the scheduler
[
https://issues.apache.org/jira/browse/AURORA-1780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15689949#comment-15689949
]
Stephan Erb commented on AURORA-1780:
-
Fix is on master. Thanks!
{code}
commit 4797dfe33ba08183fa9596a46ac8be51a64e08bb
Author: Renan DelValle
Date: Wed Nov 23 13:08:51 2016 +0100
Filter out calls to fromResource for resources that Aurora does not support
yet to avoid crashing
Added filters whenever fromResource is called for a Protos.Resource in
order to avoid Aurora crashing.
Previously only bagFromMesosResources was using the SUPPORTED_RESOURCE
filter.
Reviewed at https://reviews.apache.org/r/53923/
src/main/java/org/apache/aurora/scheduler/resources/ResourceManager.java | 23
+--
1 file changed, 17 insertions(+), 6 deletions(-)
{code}
> Offers with unknown resources types to Aurora crash the scheduler
> -
>
> Key: AURORA-1780
> URL: https://issues.apache.org/jira/browse/AURORA-1780
> Project: Aurora
> Issue Type: Bug
> Environment: vagrant
>Reporter: Renan DelValle
>Assignee: Renan DelValle
> Fix For: 0.17.0
>
>
> Taking offers from Agents which have resources that are not known to Aurora
> cause the Scheduler to crash.
> Steps to reproduce:
> {code}
> vagrant up
> sudo service mesos-slave stop
> echo
> "cpus(aurora-role):0.5;cpus(*):3.5;mem(aurora-role):1024;disk:2;gpus(*):4;test:200"
> | sudo tee /etc/mesos-slave/resources
> sudo rm -f /var/lib/mesos/meta/slaves/latest
> sudo service mesos-slave start
> {code}
> Wait around a few moments for the offer to be made to Aurora
> {code}
> I0922 02:41:57.839 [Thread-19, MesosSchedulerImpl:142] Received notification
> of lost agent: value: "cadaf569-171d-42fc-a417-fbd608ea5bab-S0"
> I0922 02:42:30.585597 2999 log.cpp:577] Attempting to append 109 bytes to
> the log
> I0922 02:42:30.585654 2999 coordinator.cpp:348] Coordinator attempting to
> write APPEND action at position 4
> I0922 02:42:30.585747 2999 replica.cpp:537] Replica received write request
> for position 4 from (10)@192.168.33.7:8083
> I0922 02:42:30.586858 2999 leveldb.cpp:341] Persisting action (125 bytes) to
> leveldb took 1.086601ms
> I0922 02:42:30.586897 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587020 2999 replica.cpp:691] Replica received learned notice
> for position 4 from @0.0.0.0:0
> I0922 02:42:30.587785 2999 leveldb.cpp:341] Persisting action (127 bytes) to
> leveldb took 746999ns
> I0922 02:42:30.587805 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587811 2999 replica.cpp:697] Replica learned APPEND action at
> position 4
> I0922 02:42:30.601 [SchedulerImpl-0, OfferManager$OfferManagerImpl:185]
> Returning offers for cadaf569-171d-42fc-a417-fbd608ea5bab-S1 for compaction.
> Sep 22, 2016 2:42:38 AM
> com.google.common.util.concurrent.ServiceManager$ServiceListener failed
> SEVERE: Service SlotSizeCounterService [FAILED] has failed in the RUNNING
> state.
> java.lang.NullPointerException: Unknown Mesos resource: name: "test"
> type: SCALAR
> scalar {
> value: 200.0
> }
> role: "*"
> at java.util.Objects.requireNonNull(Objects.java:228)
> at
> org.apache.aurora.scheduler.resources.ResourceType.fromResource(ResourceType.java:355)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.lambda$static$0(ResourceManager.java:52)
> at com.google.common.collect.Iterators$7.computeNext(Iterators.java:675)
> at
> com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:143)
> at
> com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:138)
> at java.util.Iterator.forEachRemaining(Iterator.java:115)
> at
> java.util.Spliterators$IteratorSpliterator.forEachRemaining(Spliterators.java:1801)
> at java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:481)
> at
> java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:471)
> at
> java.util.stream.ReduceOps$ReduceOp.evaluateSequential(ReduceOps.java:708)
> at java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
> at
> java.util.stream.ReferencePipeline.collect(ReferencePipeline.java:499)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromResources(ResourceManager.java:274)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromMesosResources(ResourceManager.java:239)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$OfferAdapter.get(AsyncStatsModule.java:153)
> at
> org.apache.aurora.scheduler.stats.SlotSizeCounter.run(SlotSizeCounter.java:168)
> at
>
[jira] [Commented] (AURORA-1780) Offers with unknown resources types to Aurora crash the scheduler
[
https://issues.apache.org/jira/browse/AURORA-1780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15678503#comment-15678503
]
Renan DelValle commented on AURORA-1780:
Second stab at this:
https://reviews.apache.org/r/53923/
I think this time I managed to take care of the corner cases where fromResource
gets called for Protos.Resource. The error before was due to the the filters
that called fromResource() in it (such as the NON_REVOCABLE) being placed on
before the SUPPORTED_RESOURCE filter. fromResource() was then called before the
SUPPORTED_RESOURCE filter had a change to filter out unsupported resources.
So it went something like this Iterales.Filter(Iterables.Filter(resources,
NON_REVOCABLE), SUPPORTED_RESOURCE), allowing the first filter to call
fromResource on an unknown resource and crash the scheduler.
> Offers with unknown resources types to Aurora crash the scheduler
> -
>
> Key: AURORA-1780
> URL: https://issues.apache.org/jira/browse/AURORA-1780
> Project: Aurora
> Issue Type: Bug
> Environment: vagrant
>Reporter: Renan DelValle
>Assignee: Renan DelValle
> Fix For: 0.17.0
>
>
> Taking offers from Agents which have resources that are not known to Aurora
> cause the Scheduler to crash.
> Steps to reproduce:
> {code}
> vagrant up
> sudo service mesos-slave stop
> echo
> "cpus(aurora-role):0.5;cpus(*):3.5;mem(aurora-role):1024;disk:2;gpus(*):4;test:200"
> | sudo tee /etc/mesos-slave/resources
> sudo rm -f /var/lib/mesos/meta/slaves/latest
> sudo service mesos-slave start
> {code}
> Wait around a few moments for the offer to be made to Aurora
> {code}
> I0922 02:41:57.839 [Thread-19, MesosSchedulerImpl:142] Received notification
> of lost agent: value: "cadaf569-171d-42fc-a417-fbd608ea5bab-S0"
> I0922 02:42:30.585597 2999 log.cpp:577] Attempting to append 109 bytes to
> the log
> I0922 02:42:30.585654 2999 coordinator.cpp:348] Coordinator attempting to
> write APPEND action at position 4
> I0922 02:42:30.585747 2999 replica.cpp:537] Replica received write request
> for position 4 from (10)@192.168.33.7:8083
> I0922 02:42:30.586858 2999 leveldb.cpp:341] Persisting action (125 bytes) to
> leveldb took 1.086601ms
> I0922 02:42:30.586897 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587020 2999 replica.cpp:691] Replica received learned notice
> for position 4 from @0.0.0.0:0
> I0922 02:42:30.587785 2999 leveldb.cpp:341] Persisting action (127 bytes) to
> leveldb took 746999ns
> I0922 02:42:30.587805 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587811 2999 replica.cpp:697] Replica learned APPEND action at
> position 4
> I0922 02:42:30.601 [SchedulerImpl-0, OfferManager$OfferManagerImpl:185]
> Returning offers for cadaf569-171d-42fc-a417-fbd608ea5bab-S1 for compaction.
> Sep 22, 2016 2:42:38 AM
> com.google.common.util.concurrent.ServiceManager$ServiceListener failed
> SEVERE: Service SlotSizeCounterService [FAILED] has failed in the RUNNING
> state.
> java.lang.NullPointerException: Unknown Mesos resource: name: "test"
> type: SCALAR
> scalar {
> value: 200.0
> }
> role: "*"
> at java.util.Objects.requireNonNull(Objects.java:228)
> at
> org.apache.aurora.scheduler.resources.ResourceType.fromResource(ResourceType.java:355)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.lambda$static$0(ResourceManager.java:52)
> at com.google.common.collect.Iterators$7.computeNext(Iterators.java:675)
> at
> com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:143)
> at
> com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:138)
> at java.util.Iterator.forEachRemaining(Iterator.java:115)
> at
> java.util.Spliterators$IteratorSpliterator.forEachRemaining(Spliterators.java:1801)
> at java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:481)
> at
> java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:471)
> at
> java.util.stream.ReduceOps$ReduceOp.evaluateSequential(ReduceOps.java:708)
> at java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
> at
> java.util.stream.ReferencePipeline.collect(ReferencePipeline.java:499)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromResources(ResourceManager.java:274)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromMesosResources(ResourceManager.java:239)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$OfferAdapter.get(AsyncStatsModule.java:153)
> at
> org.apache.aurora.scheduler.stats.SlotSizeCounter.run(SlotSizeCounter.java:168)
> at
>
[jira] [Commented] (AURORA-1780) Offers with unknown resources types to Aurora crash the scheduler
[
https://issues.apache.org/jira/browse/AURORA-1780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15672216#comment-15672216
]
Renan DelValle commented on AURORA-1780:
Review request available: https://reviews.apache.org/r/53831/
Would like some feedback on this approach. It seemed the best way to address
this ticket without going overboard as support for arbitrary resources is
somewhere in the pipeline (AURORA-1328). I'm open to other ways of tackling
this issue.
> Offers with unknown resources types to Aurora crash the scheduler
> -
>
> Key: AURORA-1780
> URL: https://issues.apache.org/jira/browse/AURORA-1780
> Project: Aurora
> Issue Type: Bug
> Environment: vagrant
>Reporter: Renan DelValle
>Assignee: Renan DelValle
>
> Taking offers from Agents which have resources that are not known to Aurora
> cause the Scheduler to crash.
> Steps to reproduce:
> {code}
> vagrant up
> sudo service mesos-slave stop
> echo
> "cpus(aurora-role):0.5;cpus(*):3.5;mem(aurora-role):1024;disk:2;gpus(*):4;test:200"
> | sudo tee /etc/mesos-slave/resources
> sudo rm -f /var/lib/mesos/meta/slaves/latest
> sudo service mesos-slave start
> {code}
> Wait around a few moments for the offer to be made to Aurora
> {code}
> I0922 02:41:57.839 [Thread-19, MesosSchedulerImpl:142] Received notification
> of lost agent: value: "cadaf569-171d-42fc-a417-fbd608ea5bab-S0"
> I0922 02:42:30.585597 2999 log.cpp:577] Attempting to append 109 bytes to
> the log
> I0922 02:42:30.585654 2999 coordinator.cpp:348] Coordinator attempting to
> write APPEND action at position 4
> I0922 02:42:30.585747 2999 replica.cpp:537] Replica received write request
> for position 4 from (10)@192.168.33.7:8083
> I0922 02:42:30.586858 2999 leveldb.cpp:341] Persisting action (125 bytes) to
> leveldb took 1.086601ms
> I0922 02:42:30.586897 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587020 2999 replica.cpp:691] Replica received learned notice
> for position 4 from @0.0.0.0:0
> I0922 02:42:30.587785 2999 leveldb.cpp:341] Persisting action (127 bytes) to
> leveldb took 746999ns
> I0922 02:42:30.587805 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587811 2999 replica.cpp:697] Replica learned APPEND action at
> position 4
> I0922 02:42:30.601 [SchedulerImpl-0, OfferManager$OfferManagerImpl:185]
> Returning offers for cadaf569-171d-42fc-a417-fbd608ea5bab-S1 for compaction.
> Sep 22, 2016 2:42:38 AM
> com.google.common.util.concurrent.ServiceManager$ServiceListener failed
> SEVERE: Service SlotSizeCounterService [FAILED] has failed in the RUNNING
> state.
> java.lang.NullPointerException: Unknown Mesos resource: name: "test"
> type: SCALAR
> scalar {
> value: 200.0
> }
> role: "*"
> at java.util.Objects.requireNonNull(Objects.java:228)
> at
> org.apache.aurora.scheduler.resources.ResourceType.fromResource(ResourceType.java:355)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.lambda$static$0(ResourceManager.java:52)
> at com.google.common.collect.Iterators$7.computeNext(Iterators.java:675)
> at
> com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:143)
> at
> com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:138)
> at java.util.Iterator.forEachRemaining(Iterator.java:115)
> at
> java.util.Spliterators$IteratorSpliterator.forEachRemaining(Spliterators.java:1801)
> at java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:481)
> at
> java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:471)
> at
> java.util.stream.ReduceOps$ReduceOp.evaluateSequential(ReduceOps.java:708)
> at java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
> at
> java.util.stream.ReferencePipeline.collect(ReferencePipeline.java:499)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromResources(ResourceManager.java:274)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromMesosResources(ResourceManager.java:239)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$OfferAdapter.get(AsyncStatsModule.java:153)
> at
> org.apache.aurora.scheduler.stats.SlotSizeCounter.run(SlotSizeCounter.java:168)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$SlotSizeCounterService.runOneIteration(AsyncStatsModule.java:130)
> at
> com.google.common.util.concurrent.AbstractScheduledService$ServiceDelegate$Task.run(AbstractScheduledService.java:189)
> at com.google.common.util.concurrent.Callables$3.run(Callables.java:100)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at
[jira] [Commented] (AURORA-1780) Offers with unknown resources types to Aurora crash the scheduler
[
https://issues.apache.org/jira/browse/AURORA-1780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15636746#comment-15636746
]
Joshua Cohen commented on AURORA-1780:
--
+1 sounds like the most reasonable course of action.
> Offers with unknown resources types to Aurora crash the scheduler
> -
>
> Key: AURORA-1780
> URL: https://issues.apache.org/jira/browse/AURORA-1780
> Project: Aurora
> Issue Type: Bug
> Environment: vagrant
>Reporter: Renan DelValle
>Assignee: Renan DelValle
>
> Taking offers from Agents which have resources that are not known to Aurora
> cause the Scheduler to crash.
> Steps to reproduce:
> {code}
> vagrant up
> sudo service mesos-slave stop
> echo
> "cpus(aurora-role):0.5;cpus(*):3.5;mem(aurora-role):1024;disk:2;gpus(*):4;test:200"
> | sudo tee /etc/mesos-slave/resources
> sudo rm -f /var/lib/mesos/meta/slaves/latest
> sudo service mesos-slave start
> {code}
> Wait around a few moments for the offer to be made to Aurora
> {code}
> I0922 02:41:57.839 [Thread-19, MesosSchedulerImpl:142] Received notification
> of lost agent: value: "cadaf569-171d-42fc-a417-fbd608ea5bab-S0"
> I0922 02:42:30.585597 2999 log.cpp:577] Attempting to append 109 bytes to
> the log
> I0922 02:42:30.585654 2999 coordinator.cpp:348] Coordinator attempting to
> write APPEND action at position 4
> I0922 02:42:30.585747 2999 replica.cpp:537] Replica received write request
> for position 4 from (10)@192.168.33.7:8083
> I0922 02:42:30.586858 2999 leveldb.cpp:341] Persisting action (125 bytes) to
> leveldb took 1.086601ms
> I0922 02:42:30.586897 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587020 2999 replica.cpp:691] Replica received learned notice
> for position 4 from @0.0.0.0:0
> I0922 02:42:30.587785 2999 leveldb.cpp:341] Persisting action (127 bytes) to
> leveldb took 746999ns
> I0922 02:42:30.587805 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587811 2999 replica.cpp:697] Replica learned APPEND action at
> position 4
> I0922 02:42:30.601 [SchedulerImpl-0, OfferManager$OfferManagerImpl:185]
> Returning offers for cadaf569-171d-42fc-a417-fbd608ea5bab-S1 for compaction.
> Sep 22, 2016 2:42:38 AM
> com.google.common.util.concurrent.ServiceManager$ServiceListener failed
> SEVERE: Service SlotSizeCounterService [FAILED] has failed in the RUNNING
> state.
> java.lang.NullPointerException: Unknown Mesos resource: name: "test"
> type: SCALAR
> scalar {
> value: 200.0
> }
> role: "*"
> at java.util.Objects.requireNonNull(Objects.java:228)
> at
> org.apache.aurora.scheduler.resources.ResourceType.fromResource(ResourceType.java:355)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.lambda$static$0(ResourceManager.java:52)
> at com.google.common.collect.Iterators$7.computeNext(Iterators.java:675)
> at
> com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:143)
> at
> com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:138)
> at java.util.Iterator.forEachRemaining(Iterator.java:115)
> at
> java.util.Spliterators$IteratorSpliterator.forEachRemaining(Spliterators.java:1801)
> at java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:481)
> at
> java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:471)
> at
> java.util.stream.ReduceOps$ReduceOp.evaluateSequential(ReduceOps.java:708)
> at java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
> at
> java.util.stream.ReferencePipeline.collect(ReferencePipeline.java:499)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromResources(ResourceManager.java:274)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromMesosResources(ResourceManager.java:239)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$OfferAdapter.get(AsyncStatsModule.java:153)
> at
> org.apache.aurora.scheduler.stats.SlotSizeCounter.run(SlotSizeCounter.java:168)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$SlotSizeCounterService.runOneIteration(AsyncStatsModule.java:130)
> at
> com.google.common.util.concurrent.AbstractScheduledService$ServiceDelegate$Task.run(AbstractScheduledService.java:189)
> at com.google.common.util.concurrent.Callables$3.run(Callables.java:100)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
> at
>
[jira] [Commented] (AURORA-1780) Offers with unknown resources types to Aurora crash the scheduler
[
https://issues.apache.org/jira/browse/AURORA-1780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15634516#comment-15634516
]
Stephan Erb commented on AURORA-1780:
-
Sounds good for me as well.
> Offers with unknown resources types to Aurora crash the scheduler
> -
>
> Key: AURORA-1780
> URL: https://issues.apache.org/jira/browse/AURORA-1780
> Project: Aurora
> Issue Type: Bug
> Environment: vagrant
>Reporter: Renan DelValle
>
> Taking offers from Agents which have resources that are not known to Aurora
> cause the Scheduler to crash.
> Steps to reproduce:
> {code}
> vagrant up
> sudo service mesos-slave stop
> echo
> "cpus(aurora-role):0.5;cpus(*):3.5;mem(aurora-role):1024;disk:2;gpus(*):4;test:200"
> | sudo tee /etc/mesos-slave/resources
> sudo rm -f /var/lib/mesos/meta/slaves/latest
> sudo service mesos-slave start
> {code}
> Wait around a few moments for the offer to be made to Aurora
> {code}
> I0922 02:41:57.839 [Thread-19, MesosSchedulerImpl:142] Received notification
> of lost agent: value: "cadaf569-171d-42fc-a417-fbd608ea5bab-S0"
> I0922 02:42:30.585597 2999 log.cpp:577] Attempting to append 109 bytes to
> the log
> I0922 02:42:30.585654 2999 coordinator.cpp:348] Coordinator attempting to
> write APPEND action at position 4
> I0922 02:42:30.585747 2999 replica.cpp:537] Replica received write request
> for position 4 from (10)@192.168.33.7:8083
> I0922 02:42:30.586858 2999 leveldb.cpp:341] Persisting action (125 bytes) to
> leveldb took 1.086601ms
> I0922 02:42:30.586897 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587020 2999 replica.cpp:691] Replica received learned notice
> for position 4 from @0.0.0.0:0
> I0922 02:42:30.587785 2999 leveldb.cpp:341] Persisting action (127 bytes) to
> leveldb took 746999ns
> I0922 02:42:30.587805 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587811 2999 replica.cpp:697] Replica learned APPEND action at
> position 4
> I0922 02:42:30.601 [SchedulerImpl-0, OfferManager$OfferManagerImpl:185]
> Returning offers for cadaf569-171d-42fc-a417-fbd608ea5bab-S1 for compaction.
> Sep 22, 2016 2:42:38 AM
> com.google.common.util.concurrent.ServiceManager$ServiceListener failed
> SEVERE: Service SlotSizeCounterService [FAILED] has failed in the RUNNING
> state.
> java.lang.NullPointerException: Unknown Mesos resource: name: "test"
> type: SCALAR
> scalar {
> value: 200.0
> }
> role: "*"
> at java.util.Objects.requireNonNull(Objects.java:228)
> at
> org.apache.aurora.scheduler.resources.ResourceType.fromResource(ResourceType.java:355)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.lambda$static$0(ResourceManager.java:52)
> at com.google.common.collect.Iterators$7.computeNext(Iterators.java:675)
> at
> com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:143)
> at
> com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:138)
> at java.util.Iterator.forEachRemaining(Iterator.java:115)
> at
> java.util.Spliterators$IteratorSpliterator.forEachRemaining(Spliterators.java:1801)
> at java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:481)
> at
> java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:471)
> at
> java.util.stream.ReduceOps$ReduceOp.evaluateSequential(ReduceOps.java:708)
> at java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
> at
> java.util.stream.ReferencePipeline.collect(ReferencePipeline.java:499)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromResources(ResourceManager.java:274)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromMesosResources(ResourceManager.java:239)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$OfferAdapter.get(AsyncStatsModule.java:153)
> at
> org.apache.aurora.scheduler.stats.SlotSizeCounter.run(SlotSizeCounter.java:168)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$SlotSizeCounterService.runOneIteration(AsyncStatsModule.java:130)
> at
> com.google.common.util.concurrent.AbstractScheduledService$ServiceDelegate$Task.run(AbstractScheduledService.java:189)
> at com.google.common.util.concurrent.Callables$3.run(Callables.java:100)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
> at
>
[jira] [Commented] (AURORA-1780) Offers with unknown resources types to Aurora crash the scheduler
[
https://issues.apache.org/jira/browse/AURORA-1780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15634508#comment-15634508
]
Zameer Manji commented on AURORA-1780:
--
Yes, that is the most desirable course of action for now.
> Offers with unknown resources types to Aurora crash the scheduler
> -
>
> Key: AURORA-1780
> URL: https://issues.apache.org/jira/browse/AURORA-1780
> Project: Aurora
> Issue Type: Bug
> Environment: vagrant
>Reporter: Renan DelValle
>
> Taking offers from Agents which have resources that are not known to Aurora
> cause the Scheduler to crash.
> Steps to reproduce:
> {code}
> vagrant up
> sudo service mesos-slave stop
> echo
> "cpus(aurora-role):0.5;cpus(*):3.5;mem(aurora-role):1024;disk:2;gpus(*):4;test:200"
> | sudo tee /etc/mesos-slave/resources
> sudo rm -f /var/lib/mesos/meta/slaves/latest
> sudo service mesos-slave start
> {code}
> Wait around a few moments for the offer to be made to Aurora
> {code}
> I0922 02:41:57.839 [Thread-19, MesosSchedulerImpl:142] Received notification
> of lost agent: value: "cadaf569-171d-42fc-a417-fbd608ea5bab-S0"
> I0922 02:42:30.585597 2999 log.cpp:577] Attempting to append 109 bytes to
> the log
> I0922 02:42:30.585654 2999 coordinator.cpp:348] Coordinator attempting to
> write APPEND action at position 4
> I0922 02:42:30.585747 2999 replica.cpp:537] Replica received write request
> for position 4 from (10)@192.168.33.7:8083
> I0922 02:42:30.586858 2999 leveldb.cpp:341] Persisting action (125 bytes) to
> leveldb took 1.086601ms
> I0922 02:42:30.586897 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587020 2999 replica.cpp:691] Replica received learned notice
> for position 4 from @0.0.0.0:0
> I0922 02:42:30.587785 2999 leveldb.cpp:341] Persisting action (127 bytes) to
> leveldb took 746999ns
> I0922 02:42:30.587805 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587811 2999 replica.cpp:697] Replica learned APPEND action at
> position 4
> I0922 02:42:30.601 [SchedulerImpl-0, OfferManager$OfferManagerImpl:185]
> Returning offers for cadaf569-171d-42fc-a417-fbd608ea5bab-S1 for compaction.
> Sep 22, 2016 2:42:38 AM
> com.google.common.util.concurrent.ServiceManager$ServiceListener failed
> SEVERE: Service SlotSizeCounterService [FAILED] has failed in the RUNNING
> state.
> java.lang.NullPointerException: Unknown Mesos resource: name: "test"
> type: SCALAR
> scalar {
> value: 200.0
> }
> role: "*"
> at java.util.Objects.requireNonNull(Objects.java:228)
> at
> org.apache.aurora.scheduler.resources.ResourceType.fromResource(ResourceType.java:355)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.lambda$static$0(ResourceManager.java:52)
> at com.google.common.collect.Iterators$7.computeNext(Iterators.java:675)
> at
> com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:143)
> at
> com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:138)
> at java.util.Iterator.forEachRemaining(Iterator.java:115)
> at
> java.util.Spliterators$IteratorSpliterator.forEachRemaining(Spliterators.java:1801)
> at java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:481)
> at
> java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:471)
> at
> java.util.stream.ReduceOps$ReduceOp.evaluateSequential(ReduceOps.java:708)
> at java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
> at
> java.util.stream.ReferencePipeline.collect(ReferencePipeline.java:499)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromResources(ResourceManager.java:274)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromMesosResources(ResourceManager.java:239)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$OfferAdapter.get(AsyncStatsModule.java:153)
> at
> org.apache.aurora.scheduler.stats.SlotSizeCounter.run(SlotSizeCounter.java:168)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$SlotSizeCounterService.runOneIteration(AsyncStatsModule.java:130)
> at
> com.google.common.util.concurrent.AbstractScheduledService$ServiceDelegate$Task.run(AbstractScheduledService.java:189)
> at com.google.common.util.concurrent.Callables$3.run(Callables.java:100)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
>
[jira] [Commented] (AURORA-1780) Offers with unknown resources types to Aurora crash the scheduler
[
https://issues.apache.org/jira/browse/AURORA-1780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15634485#comment-15634485
]
Renan DelValle commented on AURORA-1780:
Would everyone be OK with ignoring unknown resource types and letting the
scheduler carry on for now?
> Offers with unknown resources types to Aurora crash the scheduler
> -
>
> Key: AURORA-1780
> URL: https://issues.apache.org/jira/browse/AURORA-1780
> Project: Aurora
> Issue Type: Bug
> Environment: vagrant
>Reporter: Renan DelValle
>
> Taking offers from Agents which have resources that are not known to Aurora
> cause the Scheduler to crash.
> Steps to reproduce:
> {code}
> vagrant up
> sudo service mesos-slave stop
> echo
> "cpus(aurora-role):0.5;cpus(*):3.5;mem(aurora-role):1024;disk:2;gpus(*):4;test:200"
> | sudo tee /etc/mesos-slave/resources
> sudo rm -f /var/lib/mesos/meta/slaves/latest
> sudo service mesos-slave start
> {code}
> Wait around a few moments for the offer to be made to Aurora
> {code}
> I0922 02:41:57.839 [Thread-19, MesosSchedulerImpl:142] Received notification
> of lost agent: value: "cadaf569-171d-42fc-a417-fbd608ea5bab-S0"
> I0922 02:42:30.585597 2999 log.cpp:577] Attempting to append 109 bytes to
> the log
> I0922 02:42:30.585654 2999 coordinator.cpp:348] Coordinator attempting to
> write APPEND action at position 4
> I0922 02:42:30.585747 2999 replica.cpp:537] Replica received write request
> for position 4 from (10)@192.168.33.7:8083
> I0922 02:42:30.586858 2999 leveldb.cpp:341] Persisting action (125 bytes) to
> leveldb took 1.086601ms
> I0922 02:42:30.586897 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587020 2999 replica.cpp:691] Replica received learned notice
> for position 4 from @0.0.0.0:0
> I0922 02:42:30.587785 2999 leveldb.cpp:341] Persisting action (127 bytes) to
> leveldb took 746999ns
> I0922 02:42:30.587805 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587811 2999 replica.cpp:697] Replica learned APPEND action at
> position 4
> I0922 02:42:30.601 [SchedulerImpl-0, OfferManager$OfferManagerImpl:185]
> Returning offers for cadaf569-171d-42fc-a417-fbd608ea5bab-S1 for compaction.
> Sep 22, 2016 2:42:38 AM
> com.google.common.util.concurrent.ServiceManager$ServiceListener failed
> SEVERE: Service SlotSizeCounterService [FAILED] has failed in the RUNNING
> state.
> java.lang.NullPointerException: Unknown Mesos resource: name: "test"
> type: SCALAR
> scalar {
> value: 200.0
> }
> role: "*"
> at java.util.Objects.requireNonNull(Objects.java:228)
> at
> org.apache.aurora.scheduler.resources.ResourceType.fromResource(ResourceType.java:355)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.lambda$static$0(ResourceManager.java:52)
> at com.google.common.collect.Iterators$7.computeNext(Iterators.java:675)
> at
> com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:143)
> at
> com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:138)
> at java.util.Iterator.forEachRemaining(Iterator.java:115)
> at
> java.util.Spliterators$IteratorSpliterator.forEachRemaining(Spliterators.java:1801)
> at java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:481)
> at
> java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:471)
> at
> java.util.stream.ReduceOps$ReduceOp.evaluateSequential(ReduceOps.java:708)
> at java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
> at
> java.util.stream.ReferencePipeline.collect(ReferencePipeline.java:499)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromResources(ResourceManager.java:274)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromMesosResources(ResourceManager.java:239)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$OfferAdapter.get(AsyncStatsModule.java:153)
> at
> org.apache.aurora.scheduler.stats.SlotSizeCounter.run(SlotSizeCounter.java:168)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$SlotSizeCounterService.runOneIteration(AsyncStatsModule.java:130)
> at
> com.google.common.util.concurrent.AbstractScheduledService$ServiceDelegate$Task.run(AbstractScheduledService.java:189)
> at com.google.common.util.concurrent.Callables$3.run(Callables.java:100)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
> at
>
[jira] [Commented] (AURORA-1780) Offers with unknown resources types to Aurora crash the scheduler
[
https://issues.apache.org/jira/browse/AURORA-1780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15514331#comment-15514331
]
Renan DelValle commented on AURORA-1780:
FWIW, I have another framework running that relies on arbitrary resources
(research oriented). To run Aurora on our cluster I have to shut it all down,
remove the arbitrary resources, and bring the cluster back up. And then do the
reverse when I run my research framework. So, all in all, this issue is a
pretty big thorn on my side. On systems running systemd (tested on ubuntu
xenial) this an even nastier issue because they system brings it back up after
it crashes, hiding the issue in plain sight until the logs are checked.
> Offers with unknown resources types to Aurora crash the scheduler
> -
>
> Key: AURORA-1780
> URL: https://issues.apache.org/jira/browse/AURORA-1780
> Project: Aurora
> Issue Type: Bug
> Environment: vagrant
>Reporter: Renan DelValle
>
> Taking offers from Agents which have resources that are not known to Aurora
> cause the Scheduler to crash.
> Steps to reproduce:
> {code}
> vagrant up
> sudo service mesos-slave stop
> echo
> "cpus(aurora-role):0.5;cpus(*):3.5;mem(aurora-role):1024;disk:2;gpus(*):4;test:200"
> | sudo tee /etc/mesos-slave/resources
> sudo rm -f /var/lib/mesos/meta/slaves/latest
> sudo service mesos-slave start
> {code}
> Wait around a few moments for the offer to be made to Aurora
> {code}
> I0922 02:41:57.839 [Thread-19, MesosSchedulerImpl:142] Received notification
> of lost agent: value: "cadaf569-171d-42fc-a417-fbd608ea5bab-S0"
> I0922 02:42:30.585597 2999 log.cpp:577] Attempting to append 109 bytes to
> the log
> I0922 02:42:30.585654 2999 coordinator.cpp:348] Coordinator attempting to
> write APPEND action at position 4
> I0922 02:42:30.585747 2999 replica.cpp:537] Replica received write request
> for position 4 from (10)@192.168.33.7:8083
> I0922 02:42:30.586858 2999 leveldb.cpp:341] Persisting action (125 bytes) to
> leveldb took 1.086601ms
> I0922 02:42:30.586897 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587020 2999 replica.cpp:691] Replica received learned notice
> for position 4 from @0.0.0.0:0
> I0922 02:42:30.587785 2999 leveldb.cpp:341] Persisting action (127 bytes) to
> leveldb took 746999ns
> I0922 02:42:30.587805 2999 replica.cpp:712] Persisted action at 4
> I0922 02:42:30.587811 2999 replica.cpp:697] Replica learned APPEND action at
> position 4
> I0922 02:42:30.601 [SchedulerImpl-0, OfferManager$OfferManagerImpl:185]
> Returning offers for cadaf569-171d-42fc-a417-fbd608ea5bab-S1 for compaction.
> Sep 22, 2016 2:42:38 AM
> com.google.common.util.concurrent.ServiceManager$ServiceListener failed
> SEVERE: Service SlotSizeCounterService [FAILED] has failed in the RUNNING
> state.
> java.lang.NullPointerException: Unknown Mesos resource: name: "test"
> type: SCALAR
> scalar {
> value: 200.0
> }
> role: "*"
> at java.util.Objects.requireNonNull(Objects.java:228)
> at
> org.apache.aurora.scheduler.resources.ResourceType.fromResource(ResourceType.java:355)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.lambda$static$0(ResourceManager.java:52)
> at com.google.common.collect.Iterators$7.computeNext(Iterators.java:675)
> at
> com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:143)
> at
> com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:138)
> at java.util.Iterator.forEachRemaining(Iterator.java:115)
> at
> java.util.Spliterators$IteratorSpliterator.forEachRemaining(Spliterators.java:1801)
> at java.util.stream.AbstractPipeline.copyInto(AbstractPipeline.java:481)
> at
> java.util.stream.AbstractPipeline.wrapAndCopyInto(AbstractPipeline.java:471)
> at
> java.util.stream.ReduceOps$ReduceOp.evaluateSequential(ReduceOps.java:708)
> at java.util.stream.AbstractPipeline.evaluate(AbstractPipeline.java:234)
> at
> java.util.stream.ReferencePipeline.collect(ReferencePipeline.java:499)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromResources(ResourceManager.java:274)
> at
> org.apache.aurora.scheduler.resources.ResourceManager.bagFromMesosResources(ResourceManager.java:239)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$OfferAdapter.get(AsyncStatsModule.java:153)
> at
> org.apache.aurora.scheduler.stats.SlotSizeCounter.run(SlotSizeCounter.java:168)
> at
> org.apache.aurora.scheduler.stats.AsyncStatsModule$SlotSizeCounterService.runOneIteration(AsyncStatsModule.java:130)
> at
> com.google.common.util.concurrent.AbstractScheduledService$ServiceDelegate$Task.run(AbstractScheduledService.java:189)
> at
