mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Benjamin Bannier <benjamin.bann...@mesosphere.io>
Subject Re: Review Request 69938: Add resource decorator hook to implicitly allocate mandatory resources.
Date Tue, 12 Feb 2019 08:34:47 GMT


> On Feb. 12, 2019, 12:33 a.m., Mesos Reviewbot Windows wrote:
> > FAIL: Some of the unit tests failed. Please check the relevant logs.
> > 
> > Reviews applied: `['69938']`
> > 
> > Failed command: `Start-MesosCITesting`
> > 
> > All the build artifacts available at: http://dcos-win.westus2.cloudapp.azure.com/artifacts/mesos-reviewbot-testing/2875/mesos-review-69938
> > 
> > Relevant logs:
> > 
> > - [mesos-tests.log](http://dcos-win.westus2.cloudapp.azure.com/artifacts/mesos-reviewbot-testing/2875/mesos-review-69938/logs/mesos-tests.log):
> > 
> > ```
> > I0211 23:33:18.541034 38112 master.cpp:11334] Removing task 7445aba2-8f7f-4e39-81f3-2dedacf4d400
with resources cpus(allocated: *):4; mem(allocated: *):2048; disk(allocated: *):1024; ports(allocated:
*):[31000-32000] of framework 5dbd2f44-60f7-4adc-abef-b7d101aa059c-0000 on agent 5dbd2f44-60f7-4adc-abef-b7d101aa059c-S0
at slave(477)@192.10.1.6:64710 (windows-02.chtsmhjxogyevckjfayqqcnjda.xx.internal.cloudapp.net)
> > I0211 23:33:18.543037 44616 master.cpp:1269] Agent 5dbd2f44-60f7-4adc-abef-b7d101aa059c-S0
at slave(477)@192.10.1.6:64710 (windows-02.chtsmhjxogyevckjfayqqcnjda.xx.internal.cloudapp.net)
disconnected
> > I0211 23:33:18.543037 44616 master.cpp:3272] Disconnecting agent 5dbd2f44-60f7-4adc-abef-b7d101aa059c-S0
at slave(477)@192.10.1.6:64710 (windows-02.chtsmhjxogyevckjfayqqcnjda.xx.internal.cloudapp.net)
> > I0211 23:33:18.543037 44616 master.cpp:3291] Deactivating agent 5dbd2f44-60f7-4adc-abef-b7d101aa059c-S0
at slave(477)@192.10.1.6:64710 (windows-02.chtsmhjxogyevckjfayqqcnjda.xx.internal.cloudapp.net)
> > I0211 23:33:18.544024   688 hierarchical.cpp:358] Removed framework 5dbd2f44-60f7-4adc-abef-b7d101aa059c-0000
> > I0211 23:33:18.544024   688 hierarchical.cpp:793] Agent 5dbd2f44-60f7-4adc-abef-b7d101aa059c-S0
deactivated
> > I0211 23:33:18.545049 44772 containerizer.cpp:2477] Destroying container a8b74ba0-74f5-42a0-b1f3-5f78d854efae
in RUNNING state
> > I0211 23:33:18.545049 44772 containerizer.cpp:3144] Transitioning the state of container
a8b74ba0-74f5-42a0-b1f3-5f78d854efae from RUNNING to DESTROYING
> > I0211 23:33:18.546025 44772 launcher.cpp:161] Asked to destroy container a8b74ba0-74f5-42a0-b1f3-5f78d854efae
> > W0211 23:33:18.547099 44612 process.cpp:1423] Failed to recv on socket WindowsFD::Type::SOCKET=7556
to peer '192.10.1.6:50244': IO failed with error code: The specified network name is no longer
available.
> > 
> > W0211 23:33:18.54709[       OK ] IsolationFlag/MemoryIsolatorTest.ROOT_MemUsage/0
(692 ms)
> > [----------] 1 test from IsolationFlag/MemoryIsolatorTest (708 ms total)
> > 
> > [----------] Global test environment tear-down
> > [==========] 1095 tests from 104 test cases ran. (508950 ms total)
> > [  PASSED  ] 1094 tests.
> > [  FAILED  ] 1 test, listed below:
> > [  FAILED  ] DockerFetcherPluginTest.INTERNET_CURL_InvokeFetchByName
> > 
> >  1 FAILED TEST
> >   YOU HAVE 232 DISABLED TESTS
> > 
> > 9 44612 process.cpp:838] Failed to recv on socket WindowsFD::Type::SOCKET=7880 to
peer '192.10.1.6:50245': IO failed with error code: The specified network name is no longer
available.
> > 
> > I0211 23:33:18.642029 44616 containerizer.cpp:2983] Container a8b74ba0-74f5-42a0-b1f3-5f78d854efae
has exited
> > I0211 23:33:18.675004 45404 master.cpp:1109] Master terminating
> > I0211 23:33:18.676004 42772 hierarchical.cpp:644] Removed agent 5dbd2f44-60f7-4adc-abef-b7d101aa059c-S0
> > I0211 23:33:19.212002 44612 process.cpp:927] Stopped the socket accept loop
> > ```
> 
> Clement Michaud wrote:
>     Weird, can it be a randomness? Can we re-trigger the job?

Created https://issues.apache.org/jira/browse/MESOS-9566.


- Benjamin


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/69938/#review212725
-----------------------------------------------------------


On Feb. 11, 2019, 11:27 p.m., Clement Michaud wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/69938/
> -----------------------------------------------------------
> 
> (Updated Feb. 11, 2019, 11:27 p.m.)
> 
> 
> Review request for mesos and Benjamin Mahler.
> 
> 
> Bugs: MESOS-9315
>     https://issues.apache.org/jira/browse/MESOS-9315
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This commit introduces a master hook to decorate task resources in
> order to allocate a given amount of custom resource if the framework
> does not support it yet.
> 
> For instance, if one introduces a new custom resource in a cluster
> running frameworks not supporting this resource, there will be a mixed
> set of tasks consuming and not consuming this resource leading to
> isolation issues. By implementing this hook, a default amount can be
> allocated for a custom resources on behalf of the framework so that
> every tasks end up consuming this resource and Mesos can take it into
> account.
> 
> This implicit allocation of resource helps introducing a new custom
> resource in the clusters because, before this patch, all frameworks
> needed to be patched before introducing the new resource while now a
> default value can be applied for the frameworks not supporting the
> resource yet meaning the patches can be done later.
> 
> https://issues.apache.org/jira/browse/MESOS-9315
> 
> 
> Diffs
> -----
> 
>   include/mesos/hook.hpp 019887095e7845d5a65d133b0f58091d262ec55b 
>   src/examples/test_hook_module.cpp c4f449512a4cc150de8a99f44a525b96a2fc1ae2 
>   src/hook/manager.hpp b3d4f5198588068d3b28a57cffb3754b55e33b51 
>   src/hook/manager.cpp 3e71a26f8c0fcfefecc93d70f8a9d6c2d7fdcc6c 
>   src/master/master.cpp b4faf2b077a0288ba36195b7a21402932489d316 
>   src/tests/hook_tests.cpp d8aa35e0027d589044bb131b460311721bd36609 
> 
> 
> Diff: https://reviews.apache.org/r/69938/diff/3/
> 
> 
> Testing
> -------
> 
> I added a test showing that if a task was missing network_bandwidth resource in the TaskInfo,
the hook injects a default value on behalf of the framework.
> 
> 
> Thanks,
> 
> Clement Michaud
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message