[
https://issues.apache.org/jira/browse/BEAM-3645?focusedWorklogId=284002&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-284002
]
ASF GitHub Bot logged work on BEAM-3645:
----------------------------------------
Author: ASF GitHub Bot
Created on: 29/Jul/19 03:29
Start Date: 29/Jul/19 03:29
Worklog Time Spent: 10m
Work Description: Hannah-Jiang commented on issue #8979: [BEAM-3645] add
multiplexing for python FnApiRunner
URL: https://github.com/apache/beam/pull/8979#issuecomment-515830723
The reason threads are not all killed immediately for direct runners for 10%
of time is we are starting grpc servers and close it with direct runners.
Threads created for grpc servers at FnApi side are all closed immediately, but
it seems like a thread at grpc library is closed a little late. If confirmed
that if we don't start grpc servers for direct runner, then all thread are
closed immediately (sample size is 3,000.), so it would be resolved when we
don't create grpc servers for direct embedded runners.
A side note about thread name is, we can add thread_name_prefix for
ThreadPoolExecutor. This is supported by concurrent lib from v3.6. However,
both Python2 and Python3 are using [concurrent
v3.2](https://docs.python.org/3/library/concurrent.futures.html), so even if we
add it, thread names are in default format for threads created through
ThreadPoolExecutor. However, it is still a good practice to add it in case
concurrent lib version is updated in future and we will have correct thread
names for free.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 284002)
Time Spent: 33h 10m (was: 33h)
> Support multi-process execution on the FnApiRunner
> --------------------------------------------------
>
> Key: BEAM-3645
> URL: https://issues.apache.org/jira/browse/BEAM-3645
> Project: Beam
> Issue Type: Improvement
> Components: sdk-py-core
> Affects Versions: 2.2.0, 2.3.0
> Reporter: Charles Chen
> Assignee: Hannah Jiang
> Priority: Major
> Fix For: 2.15.0
>
> Time Spent: 33h 10m
> Remaining Estimate: 0h
>
> https://issues.apache.org/jira/browse/BEAM-3644 gave us a 15x performance
> gain over the previous DirectRunner. We can do even better in multi-core
> environments by supporting multi-process execution in the FnApiRunner, to
> scale past Python GIL limitations.
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
