aurora-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Kai Huang <texasred2...@hotmail.com>
Subject Re: Review Request 59699: Improve task history pruning by batch deleting tasks
Date Thu, 01 Jun 2017 22:42:49 GMT


> On June 1, 2017, 9:22 p.m., David McLaughlin wrote:
> > src/main/java/org/apache/aurora/scheduler/state/StateManagerImpl.java
> > Lines 389-391 (original), 374-376 (patched)
> > <https://reviews.apache.org/r/59699/diff/1-2/?file=1736080#file1736080line391>
> >
> >     We probably don't even need the separate events. We could just have:
> >     
> >         eventSink.post(createDeleteEvent(taskStore, taskIds));
> 
> Kai Huang wrote:
>     So this will change the semantics that: The Delete Event is published after we delete
task from TaskStore?
> 
> David McLaughlin wrote:
>     I don't think it changes the semantics at all? In both cases, the events are published
after the batch delete. All I'm suggesting is we send a single event for the batch rather
than one event per task in the batch.

addressed.


- Kai


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/59699/#review176670
-----------------------------------------------------------


On June 1, 2017, 10:41 p.m., Kai Huang wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/59699/
> -----------------------------------------------------------
> 
> (Updated June 1, 2017, 10:41 p.m.)
> 
> 
> Review request for Aurora, David McLaughlin and Santhosh Kumar.
> 
> 
> Bugs: AURORA-1929
>     https://issues.apache.org/jira/browse/AURORA-1929
> 
> 
> Repository: aurora
> 
> 
> Description
> -------
> 
> Improve task history pruning by batch deleting tasks.
> 
> The `'aurora_admin prune_tasks'` endpoint seems to be very slow when the cluster has
a large number of inactive tasks.
> 
> This CR batches all removeTasks operations and execute them all at once to avoid additional
cost of coalescing. The fix will also benefit implicit task history pruning since it has similar
underlying implementation. See https://issues.apache.org/jira/browse/AURORA-1929 for more
information and details.
> 
> 
> Diffs
> -----
> 
>   src/main/java/org/apache/aurora/scheduler/state/StateManagerImpl.java 73878210f9028901fda3b08e66c6a63c24260d35

> 
> 
> Diff: https://reviews.apache.org/r/59699/diff/3/
> 
> 
> Testing
> -------
> 
> __unit_tests:__
> 
> ./build-support/jenkins/build.sh
> 
> No unit tests were created for this patch since it does not add new functionalities or
alter the interface, but improves the efficiency of the existing code.
> 
> __e2e tests:__
> 
> Attached was a screenshot of the task history pruning benchmark obtained from a scale
test in Twitter's test cluster.
> 
> - Before applying this patch, the task history pruning takes ~30 minutes on 130K tasks.
> 
> - After applying the patch, the pruning takes ~1 minute.
> 
> 
> File Attachments
> ----------------
> 
> task_history_pruning_benchmark.png
>   https://reviews.apache.org/media/uploaded/files/2017/06/01/74eb5104-d338-4530-abd2-b82fbdc6bf84__task_history_pruning_benchmark.png
> 
> 
> Thanks,
> 
> Kai Huang
> 
>


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