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 67670: Added a unit test for disappeared profiles.
Date Wed, 20 Jun 2018 13:14:05 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/67670/#review205061
-----------------------------------------------------------




src/tests/storage_local_resource_provider_tests.cpp
Lines 1047 (patched)
<https://reviews.apache.org/r/67670/#comment288023>

    If this number depends on e.g., the allocation interval or the number of allocation cycles
we perform below, it might be worthwhile to document that. I am not sure if there's any dependency.



src/tests/storage_local_resource_provider_tests.cpp
Lines 1058 (patched)
<https://reviews.apache.org/r/67670/#comment288012>

    Let's pass this to `StartMaster`.



src/tests/storage_local_resource_provider_tests.cpp
Lines 1066 (patched)
<https://reviews.apache.org/r/67670/#comment288011>

    Why is this required? It looks like this is the only driver for making this a `ROOT` test.
If it is required we should document why; else let's make this a non-`ROOT` test.
    
    nit: We can remove the empty line below.



src/tests/storage_local_resource_provider_tests.cpp
Lines 1165-1168 (patched)
<https://reviews.apache.org/r/67670/#comment288016>

    This relies strongly on the way the master interprets a zero time, and  how the allocation
interval is related to the default value the master would currently use. Can we set a value
calculated from `masterFlags.allocation_interval` instead to decouple this?



src/tests/storage_local_resource_provider_tests.cpp
Lines 1180 (patched)
<https://reviews.apache.org/r/67670/#comment288017>

    Let's `ASSERT` here that we were offered the 4GB `disk` we expect.



src/tests/storage_local_resource_provider_tests.cpp
Lines 1204-1205 (patched)
<https://reviews.apache.org/r/67670/#comment288019>

    The point here seems to be to simulate a situation where an agent update and an operation
cross each other's paths between master and agent (and them being incompatible).
    
    Also: `s/hold/held/`



src/tests/storage_local_resource_provider_tests.cpp
Lines 1284 (patched)
<https://reviews.apache.org/r/67670/#comment288022>

    `storagePool.disk().get()`?


- Benjamin Bannier


On June 20, 2018, 7:34 a.m., Chun-Hung Hsiao wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/67670/
> -----------------------------------------------------------
> 
> (Updated June 20, 2018, 7:34 a.m.)
> 
> 
> Review request for mesos, Benjamin Bannier, Jie Yu, and Jan Schlicht.
> 
> 
> Bugs: MESOS-8995
>     https://issues.apache.org/jira/browse/MESOS-8995
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> The `ROOT_ProfileDisappeared` tests if a `CREATE_VOLUME` consuming a RAW
> disk with a disappeared profile will be dropped, and if the disk space
> freed by a `DESTROY_VOLUME` destroying a volume with a disappeared
> profile will be recovered with a newly appeared profile.
> 
> 
> Diffs
> -----
> 
>   src/tests/storage_local_resource_provider_tests.cpp 1afe8a8e0413ef225e952cd9cd6376e5d82774e5

> 
> 
> Diff: https://reviews.apache.org/r/67670/diff/1/
> 
> 
> Testing
> -------
> 
> sudo make check
> 
> 
> Thanks,
> 
> Chun-Hung Hsiao
> 
>


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