mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Neil Conway <>
Subject Re: Review Request 54999: Fixed test flakiness due to floating point conversions.
Date Wed, 28 Dec 2016 21:23:37 GMT

This is an automatically generated e-mail. To reply, visit:

(Updated Dec. 28, 2016, 9:23 p.m.)

Review request for mesos, Alexander Rukletsov and Benjamin Bannier.


Fix typo, tweak comment.

Bugs: MESOS-6837

Repository: mesos


`FaultToleranceTest.FrameworkReregister` and
`MasterTest.FailoverAgentReregisterFirst` both examined a timestamp
value returned by an HTTP endpoint. Such values are the result of
several conversions (`double` to `string` to `JSON::Value`); this might
result in the returned value being an integer one larger/smaller than we
expect. Hence, make the comparison within an epsilon of 1.

A similar issue in `SlaveTest.StateEndpoint` was fixed in

Diffs (updated)

  src/tests/fault_tolerance_tests.cpp 24747fab20f5b107b9c23a271b753e83f05bbee3 
  src/tests/master_tests.cpp 2d0cd8244ded44e76f0eee3d87327ff526db5208 
  src/tests/slave_tests.cpp 67a6aed8c66a21c94c106b52dff75cbdc41fcf69 



`make check`

Note that I wasn't able to repro the test failure on my laptop, but it was observed on CI.


Neil Conway

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