benchmark: split timers benchmark and refactor #9497

Trott · 2016-11-07T01:53:59Z

Checklist

make -j8 test (UNIX), or vcbuild test nosign (Windows) passes
tests and/or benchmarks are included
commit message follows commit guidelines

Affected core subsystem(s)

benchmark timers

Description of change

The depth benchmark for timers sets a timer that sets a timer that sets
a timer that... 500K of them.

Since each timer has to wait for the next tick of the event loop this
benchmark takes a very long time to run compared to the breadth
test that is already in the file. This may be more of an event loop
benchmark than a timer benchmark.

Reduce the number of iterations for the depth test as it's really just
running the iterations in sequence, not in parallel. And even on an
infinitely fast machine, it would take over 8 minutes to run because
each tick of the event loop would have to wait 1ms before firing the
timer.

Split the depth and breadth benchmarks so that their N values can be
set independently.

Do some minor refactoring to the benchmarks (but no ES6 additions so
that the benchmarks can still be run with old versions of Node.js).

Refs: #9493

AndreasMadsen · 2016-11-07T07:26:56Z

benchmark/timers/timers-breadth.js

+
+var bench = common.createBenchmark(main, {
+  thousands: [500],
+  type: ['breadth']


the filename is included in the output, so this is not needed.

Removed it! Thanks!

The depth benchmark for timers sets a timer that sets a timer that sets a timer that... 500K of them. Since each timer has to wait for the next tick of the event loop this benchmark takes a very long time to run compared to the breadth test that is already in the file. This may be more of an event loop benchmark than a timer benchmark. Reduce the number of iterations for the depth test as it's really just running the iterations in sequence, not in parallel. And even on an infinitely fast machine, it would take over 8 minutes to run because each tick of the event loop would have to wait 1ms before firing the timer. Split the depth and breadth benchmarks so that their `N` values can be set independently. Do some minor refactoring to the benchmarks (but no ES6 additions so that the benchmarks can still be run with old versions of Node.js). Refs: nodejs#9493

AndreasMadsen

LGTM

lpinca · 2016-11-08T08:27:39Z

benchmark/timers/timers-breadth.js

+var common = require('../common.js');
+
+var bench = common.createBenchmark(main, {
+  thousands: [500],


Nit: trailing comma.

lpinca · 2016-11-08T08:28:10Z

benchmark/timers/timers-depth.js

+var common = require('../common.js');
+
+var bench = common.createBenchmark(main, {
+  thousands: [1],


Fishrock123

eh, seems fine. Still need better/different benchmarks.

I'll try to dedicate a bit of time this week to it.

Trott · 2016-11-10T00:48:59Z

CI: https://ci.nodejs.org/job/node-test-pull-request/4815/

Trott · 2016-11-10T03:19:38Z

Landed in ef6a5cf

The depth benchmark for timers sets a timer that sets a timer that sets a timer that... 500K of them. Since each timer has to wait for the next tick of the event loop this benchmark takes a very long time to run compared to the breadth test that is already in the file. This may be more of an event loop benchmark than a timer benchmark. Reduce the number of iterations for the depth test as it's really just running the iterations in sequence, not in parallel. And even on an infinitely fast machine, it would take over 8 minutes to run because each tick of the event loop would have to wait 1ms before firing the timer. Split the depth and breadth benchmarks so that their `N` values can be set independently. Do some minor refactoring to the benchmarks (but no ES6 additions so that the benchmarks can still be run with old versions of Node.js). Refs: nodejs#9493 PR-URL: nodejs#9497 Reviewed-By: Andreas Madsen <amwebdk@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>

The depth benchmark for timers sets a timer that sets a timer that sets a timer that... 500K of them. Since each timer has to wait for the next tick of the event loop this benchmark takes a very long time to run compared to the breadth test that is already in the file. This may be more of an event loop benchmark than a timer benchmark. Reduce the number of iterations for the depth test as it's really just running the iterations in sequence, not in parallel. And even on an infinitely fast machine, it would take over 8 minutes to run because each tick of the event loop would have to wait 1ms before firing the timer. Split the depth and breadth benchmarks so that their `N` values can be set independently. Do some minor refactoring to the benchmarks (but no ES6 additions so that the benchmarks can still be run with old versions of Node.js). Refs: #9493 PR-URL: #9497 Reviewed-By: Andreas Madsen <amwebdk@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>

MylesBorins · 2016-12-13T20:02:23Z

@Trott do you think this should be backported?

Trott · 2016-12-13T22:03:57Z

@thealphanerd Yes. (Is landing cleanly for me on both v4.x-staging and v6.x-staging.)

The depth benchmark for timers sets a timer that sets a timer that sets a timer that... 500K of them. Since each timer has to wait for the next tick of the event loop this benchmark takes a very long time to run compared to the breadth test that is already in the file. This may be more of an event loop benchmark than a timer benchmark. Reduce the number of iterations for the depth test as it's really just running the iterations in sequence, not in parallel. And even on an infinitely fast machine, it would take over 8 minutes to run because each tick of the event loop would have to wait 1ms before firing the timer. Split the depth and breadth benchmarks so that their `N` values can be set independently. Do some minor refactoring to the benchmarks (but no ES6 additions so that the benchmarks can still be run with old versions of Node.js). Refs: #9493 PR-URL: #9497 Reviewed-By: Andreas Madsen <amwebdk@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>

Trott added timers Issues and PRs related to the timers subsystem / setImmediate, setInterval, setTimeout. benchmark Issues and PRs related to the benchmark subsystem. labels Nov 7, 2016

Trott mentioned this pull request Nov 7, 2016

timers "depth" benchmark usefulness #9493

Closed

AndreasMadsen reviewed Nov 7, 2016

View reviewed changes

Trott added 2 commits November 7, 2016 19:58

squash: remove type

03ebf8a

Trott force-pushed the timers-benchmark branch from 1df4ac9 to 03ebf8a Compare November 8, 2016 04:01

AndreasMadsen approved these changes Nov 8, 2016

View reviewed changes

lpinca reviewed Nov 8, 2016

View reviewed changes

cjihrig approved these changes Nov 8, 2016

View reviewed changes

Fishrock123 approved these changes Nov 8, 2016

View reviewed changes

Trott closed this Nov 10, 2016

MylesBorins added lts-watch-v4.x labels Dec 13, 2016

MylesBorins added land-on-v4.x and removed lts-watch-v4.x labels Dec 21, 2016

This was referenced Dec 21, 2016

V6.9.3 proposal #10394

Merged

V4.7.1 proposal #10395

Merged

Trott deleted the timers-benchmark branch January 13, 2022 22:44

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

benchmark: split timers benchmark and refactor #9497

benchmark: split timers benchmark and refactor #9497

Trott commented Nov 7, 2016

AndreasMadsen Nov 7, 2016

Trott Nov 8, 2016

AndreasMadsen left a comment

lpinca Nov 8, 2016

lpinca Nov 8, 2016

Fishrock123 left a comment •

edited

Loading

Trott commented Nov 10, 2016

Trott commented Nov 10, 2016

MylesBorins commented Dec 13, 2016

Trott commented Dec 13, 2016

benchmark: split timers benchmark and refactor #9497

benchmark: split timers benchmark and refactor #9497

Conversation

Trott commented Nov 7, 2016

Checklist

Affected core subsystem(s)

Description of change

AndreasMadsen Nov 7, 2016

Choose a reason for hiding this comment

Trott Nov 8, 2016

Choose a reason for hiding this comment

AndreasMadsen left a comment

Choose a reason for hiding this comment

lpinca Nov 8, 2016

Choose a reason for hiding this comment

lpinca Nov 8, 2016

Choose a reason for hiding this comment

Fishrock123 left a comment • edited Loading

Choose a reason for hiding this comment

Trott commented Nov 10, 2016

Trott commented Nov 10, 2016

MylesBorins commented Dec 13, 2016

Trott commented Dec 13, 2016

Fishrock123 left a comment •

edited

Loading