-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71316/#review217362
-----------------------------------------------------------




src/tests/master_draining_tests.cpp
Lines 178 (patched)
<https://reviews.apache.org/r/71316/#comment304652>

    In this case, we probably want to either not set the max grace period, or 
set it to something long, since we're trying to verify that an agent with 
nothing running will immediately be marked drained. While the agent won't be 
_immediately_ marked gone if a task is running and the max grace period is set 
to zero, it would still happen very quickly.



src/tests/master_draining_tests.cpp
Lines 222-250 (patched)
<https://reviews.apache.org/r/71316/#comment304653>

    We probably don't need to verify the state output of all 3 endpoints in 
both versions of this test, I think just one will do. WDYT?



src/tests/master_draining_tests.cpp
Lines 281 (patched)
<https://reviews.apache.org/r/71316/#comment304654>

    This test is so short now :heart_eyes:



src/tests/master_draining_tests.cpp
Lines 313-325 (patched)
<https://reviews.apache.org/r/71316/#comment304656>

    Can we do something to verify that the agent has been marked unreachable 
here? Either specifying the type of the registry operation in the EXPECT_CALL, 
or inspecting some API output afterward?



src/tests/master_draining_tests.cpp
Lines 360-361 (patched)
<https://reviews.apache.org/r/71316/#comment304658>

    Can we also verify via master API output that the agent is DRAINED after it 
reregisters?


- Greg Mann


On Aug. 19, 2019, 9:57 p.m., Joseph Wu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71316/
> -----------------------------------------------------------
> 
> (Updated Aug. 19, 2019, 9:57 p.m.)
> 
> 
> Review request for mesos, Benno Evers and Greg Mann.
> 
> 
> Bugs: MESOS-9892
>     https://issues.apache.org/jira/browse/MESOS-9892
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This splits the existing agent draining tests into two variants:
> 1) where the agent has nothing running, and
> 2) where the agent has one task running.
> 
> 
> Diffs
> -----
> 
>   src/tests/master_draining_tests.cpp PRE-CREATION 
> 
> 
> Diff: https://reviews.apache.org/r/71316/diff/1/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Joseph Wu
> 
>

Reply via email to