[
https://issues.apache.org/jira/browse/BEAM-8474?focusedWorklogId=335635&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-335635
]
ASF GitHub Bot logged work on BEAM-8474:
----------------------------------------
Author: ASF GitHub Bot
Created on: 29/Oct/19 17:30
Start Date: 29/Oct/19 17:30
Worklog Time Spent: 10m
Work Description: boyuanzz commented on issue #9891: [BEAM-8474] A
microbenchmark for Python FnApiRunner:
URL: https://github.com/apache/beam/pull/9891#issuecomment-547540201
> Results:
>
> ```
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 1 of
10, per element time cost: 3.6778 sec
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 2 of
10, per element time cost: 0.053498 sec
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 3 of
10, per element time cost: 0.0299434 sec
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 4 of
10, per element time cost: 0.0211154 sec
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 5 of
10, per element time cost: 0.0170031 sec
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 6 of
10, per element time cost: 0.0150809 sec
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 7 of
10, per element time cost: 0.013218 sec
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 8 of
10, per element time cost: 0.0119685 sec
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 9 of
10, per element time cost: 0.0107382 sec
> run_single_pipeline, 1 element(s) at start, 100 growth per run: run 10 of
10, per element time cost: 0.0103208 sec
>
>
> Fixed cost 4.537164939085642
> Per-element 0.005474923321695039
> R^2 0.9518956932309706
> ```
Changes look good to me. I would attach this perf result and suggested
regression rate into fn_api_runner_microbenchmark.py.
Would you mind explaining a little bit more on choosing `Create`,
`GroupByKey` and `Pardo` with `Timer` to construct pipeline? Which particular
piece of FnApiRunner that you want to benchmark?
----------------------------------------------------------------
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: 335635)
Time Spent: 2h (was: 1h 50m)
> A microbenchmark that exercises the FnAPI runner functionality
> --------------------------------------------------------------
>
> Key: BEAM-8474
> URL: https://issues.apache.org/jira/browse/BEAM-8474
> Project: Beam
> Issue Type: Improvement
> Components: sdk-py-core
> Reporter: Pablo Estrada
> Assignee: Pablo Estrada
> Priority: Major
> Time Spent: 2h
> Remaining Estimate: 0h
>
> The main data paths for the Fn API runner are exercised by:
> * Side inputs
> * GBK
> * State
> * Timers
> * SDF?
> A microbenchmark would have a number of stages that exercise one or more of
> these data paths.
> A microbenchmark suite may have more than one pipeline.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
