mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mesos Reviewbot Windows <revi...@mesos.apache.org>
Subject Re: Review Request 70378: Wrapped access to hashmaps in `frameworks.principals` and `authorized`.
Date Fri, 05 Apr 2019 17:36:30 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70378/#review214426
-----------------------------------------------------------



FAIL: Some of the unit tests failed. Please check the relevant logs.

Reviews applied: `['70408', '70377', '70378']`

Failed command: `Start-MesosCITesting`

All the build artifacts available at: http://dcos-win.westus2.cloudapp.azure.com/artifacts/mesos-reviewbot-testing/3096/mesos-review-70378

Relevant logs:

- [mesos-tests.log](http://dcos-win.westus2.cloudapp.azure.com/artifacts/mesos-reviewbot-testing/3096/mesos-review-70378/logs/mesos-tests.log):

```
I0405 17:36:12.889297 37920 hierarchical.cpp:392] Removed framework ec3ba3f9-4a09-4ddb-a368-5c27edcf2f5d-0000
I0405 17:36:12.890295 37920 hierarchical.cpp:829] Agent ec3ba3f9-4a09-4ddb-a368-5c27edcf2f5d-S0
deactivated
I0405 17:36:12.891297 39300 containerizer.cpp:2576] Destroying container 3281ae8e-54e2-404f-bcda-94cca52835b3
in RUNNING state
I0405 17:36:12.891297 39300 containerizer.cpp:3278] Transitioning the state of container 3281ae8e-54e2-404f-bcda-94cca52835b3
from RUNNING to DESTROYING
I0405 17:36:12.891297 39300 launcher.cpp:161] Asked to destroy container 3281ae8e-54e2-404f-bcda-94cca52835b3
W0405 17:36:12.892[       OK ] IsolationFlag/MemoryIsolatorTest.ROOT_MemUsage/0 (789 ms)
[----------] 1 test from IsolationFlag/MemoryIsolatorTest (808 ms total)

[----------] Global test environment tear-down
[==========] 1160 tests from 107 test cases ran. (570102 ms total)
[  PASSED  ] 1155 tests.
[  FAILED  ] 5 tests, listed below:
[  FAILED  ] FaultToleranceTest.ChangePrincipalOnReRegistrationFails
[  FAILED  ] DockerFetcherPluginTest.INTERNET_CURL_FetchImage
[  FAILED  ] DockerFetcherPluginTest.INTERNET_CURL_InvokeFetchByName
[  FAILED  ] ContentType/SchedulerHttpApiTest.PrincipalChangeFails/0, where GetParam() = "application/x-protobuf"
[  FAILED  ] ContentType/SchedulerHttpApiTest.PrincipalChangeFails/1, where GetParam() = "application/json"

 5 FAILED TESTS
  YOU HAVE 231 DISABLED TESTS

343 33936 process.cpp:1423] Failed to recv on socket WindowsFD::Type::SOCKET=2092 to peer
'192.10.1.7:59166': IO failed with error code: The specified network name is no longer available.

W0405 17:36:12.893293 33936 process.cpp:838] Failed to recv on socket WindowsFD::Type::SOCKET=9496
to peer '192.10.1.7:59167': IO failed with error code: The specified network name is no longer
available.

I0405 17:36:12.979327 33400 containerizer.cpp:3117] Container 3281ae8e-54e2-404f-bcda-94cca52835b3
has exited
I0405 17:36:13.011386 38712 master.cpp:1135] Master terminating
I0405 17:36:13.013305 37920 hierarchical.cpp:680] Removed agent ec3ba3f9-4a09-4ddb-a368-5c27edcf2f5d-S0
I0405 17:36:13.534353 33936 process.cpp:927] Stopped the socket accept loop
```

- Mesos Reviewbot Windows


On April 3, 2019, 9:01 a.m., Andrei Sekretenko wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70378/
> -----------------------------------------------------------
> 
> (Updated April 3, 2019, 9:01 a.m.)
> 
> 
> Review request for mesos and Gastón Kleiman.
> 
> 
> Bugs: MESOS-2842
>     https://issues.apache.org/jira/browse/MESOS-2842
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> !! This code is completely disposable and possibly not properly located; the next patch
actually does not depend on this one.
> 
> ------
> 
> After having a look at the Master code, I realized that the hashmaps `frameworks.principlas`
and `authorized` should satisfy the following property: the value corresponding to a given
key never changes.
> 
> Currently, this invariant is enforced MANUALLY with hacks of three kinds. 
> The first hack is constructs like this:
> 
> CHECK(!h.contains(key));
> h[key] = newValue;
> 
> The second hack is much worse: it is avoiding calling non-const `hashmap::operator[]()`
for nonexisting keys.
> 
> And the third one is even worse: it is avoiding this situation:
> h.erase(key);
> h[key] = newValue;
> 
> To make sure that at least the first two hacks has been used properly, I added this wrapper
around hashmap.
> Actually, this is a dirty hack itself, and not a proper decomposition of the God Object
antipattern that, IMO, has somehow started to grow inside of the Master.
> 
> Unfortunately, there is no such simple way to validate that the third kind of hacks is
used properly. (In fact, it is not - and this is the direct cause of MESOS-2842.)
> 
> 
> Diffs
> -----
> 
>   src/master/master.hpp 94891af9deeaddb3333fc9d6eabb243aed97f7b7 
>   src/master/master.cpp cf5caa0893ba1387a1f3a9d129ecd7d974f776bd 
> 
> 
> Diff: https://reviews.apache.org/r/70378/diff/1/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Andrei Sekretenko
> 
>


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