mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Joseph Wu <jos...@mesosphere.io>
Subject Re: Review Request 44258: Fixed http endpoint trigger two inverse offer calls.
Date Thu, 03 Mar 2016 20:15:05 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/44258/#review121923
-----------------------------------------------------------




src/master/http.cpp (lines 1996 - 1999)
<https://reviews.apache.org/r/44258/#comment183778>

    Instead of these two inner comments, can you instead have an overall comment like:
    ```
    // Update the unavailability for each existing machine, except for machines going from
`UP to `DOWN` (handled in the next loop).  
    ```
    
    Also, can you add a TODO here, to merge this logic with `Master::updateUnavailability`?
 (Having it in two places results in more conditionals.)



src/master/http.cpp (lines 2014 - 2016)
<https://reviews.apache.org/r/44258/#comment183780>

    These two bullet points say the same thing.


- Joseph Wu


On March 2, 2016, 10:05 p.m., Guangya Liu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/44258/
> -----------------------------------------------------------
> 
> (Updated March 2, 2016, 10:05 p.m.)
> 
> 
> Review request for mesos, Anand Mazumdar, Joris Van Remoortere, and Joseph Wu.
> 
> 
> Bugs: MESOS-4831
>     https://issues.apache.org/jira/browse/MESOS-4831
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> There is a bug when setting host maintain with http endpoint: https://github.com/apache/mesos/blob/master/src/master/http.cpp#L1987-L2021
> The logic is as this:
> 1) Get all host list from maintain window and put it to updated hashmap.
> 2) If the machine in was in updated was also in master->machines, call master updateUnavailability
to trigger recoverResources, updateUnavailability etc in allocator
> 3) Otherwise, clear the unavailabity time window for the machine.
> 4) Update each new machines in updated to call master updateUnavailability
> 
> But the logic in step 4) is getting all machines from the schedule windows but not the
machines that is new to the cluster, this caused master get two updateUnavailability calls
for a machine in the updated hashmap.
> 
> The fix is filter machines in updated hashmap when handling new machines.
> 
> 
> Diffs
> -----
> 
>   src/master/http.cpp 5e9e28e904ba0045ee27eb828f47231632a91d74 
>   src/tests/master_maintenance_tests.cpp 3faa8136cf57276295553910319480028f433e4c 
> 
> Diff: https://reviews.apache.org/r/44258/diff/
> 
> 
> Testing
> -------
> 
> make
> make check
>  ./bin/mesos-tests.sh --gtest_filter="MasterMaintenanceTest.*" --verbose
> 
> 
> Thanks,
> 
> Guangya Liu
> 
>


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