test: fix flaky test-regress-GH-897 - #10903
Closed
Trott wants to merge 1 commit into
Closed
Conversation
Copy link
Copy Markdown
Member
Sorry, something went wrong.
Even after being moved to `sequential` in 1ce05ad, `test-regress-nodejsGH-897` still was occasionally flaky on Raspberry Pi devices on CI. The test is especially sensitive to resource constraints. It failed reliably on my laptop if I moved it to `parallel` and ran 32 competing node test processes. Even for a flaky test, that's unusually low. I typically don't see problems, even for flaky tests, until I get up to around four times that number. On a Raspberry Pi, of course, that sensitivity to resource constraints will manifest much sooner. This change checks the order of timers firing, rather than the duration before a timer is fired. This eliminates the sensitivity to resource constraints. The test can now be moved back to `parallel`. I am able to run many copies of the test simultaneously without seeing test failures. Fixes: nodejs#10073
Copy link
Copy Markdown
Member
Author
|
/cc @nodejs/testing @Fishrock123 @misterdjules |
Sorry, something went wrong.
Copy link
Copy Markdown
Member
Author
Sorry, something went wrong.
Copy link
Copy Markdown
Contributor
|
If you go with this approach, do you think it might be worth throwing a 0 millisecond timeout into the mix? |
Sorry, something went wrong.
Copy link
Copy Markdown
Member
Author
I was going to do that in a subsequent PR (although if someone eager wants to beat me to it, I will be happy to not do it myself). Basically, find out if we're already testing for things like 0, -1, undefined, those sorts of things. Make sure they all get coerced to 1. If so, consolidate into a single file. If not, move them here, rename the file, and have it be the general "this makes sure your wonky values get coerced to 1 in timers" test. |
Sorry, something went wrong.
santigimeno
approved these changes
Jan 20, 2017
lpinca
approved these changes
Jan 21, 2017
Trott
added a commit
to Trott/io.js
that referenced
this pull request
Jan 23, 2017
Even after being moved to `sequential` in 1ce05ad, `test-regress-nodejsGH-897` still was occasionally flaky on Raspberry Pi devices on CI. The test is especially sensitive to resource constraints. It failed reliably on my laptop if I moved it to `parallel` and ran 32 competing node test processes. Even for a flaky test, that's unusually low. I typically don't see problems, even for flaky tests, until I get up to around four times that number. On a Raspberry Pi, of course, that sensitivity to resource constraints will manifest much sooner. This change checks the order of timers firing, rather than the duration before a timer is fired. This eliminates the sensitivity to resource constraints. The test can now be moved back to `parallel`. I am able to run many copies of the test simultaneously without seeing test failures. PR-URL: nodejs#10903 Fixes: nodejs#10073 Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com> Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
italoacasas
pushed a commit
to italoacasas/node
that referenced
this pull request
Jan 25, 2017
Even after being moved to `sequential` in 1ce05ad, `test-regress-nodejsGH-897` still was occasionally flaky on Raspberry Pi devices on CI. The test is especially sensitive to resource constraints. It failed reliably on my laptop if I moved it to `parallel` and ran 32 competing node test processes. Even for a flaky test, that's unusually low. I typically don't see problems, even for flaky tests, until I get up to around four times that number. On a Raspberry Pi, of course, that sensitivity to resource constraints will manifest much sooner. This change checks the order of timers firing, rather than the duration before a timer is fired. This eliminates the sensitivity to resource constraints. The test can now be moved back to `parallel`. I am able to run many copies of the test simultaneously without seeing test failures. PR-URL: nodejs#10903 Fixes: nodejs#10073 Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com> Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
italoacasas
pushed a commit
to italoacasas/node
that referenced
this pull request
Jan 27, 2017
Even after being moved to `sequential` in 1ce05ad, `test-regress-nodejsGH-897` still was occasionally flaky on Raspberry Pi devices on CI. The test is especially sensitive to resource constraints. It failed reliably on my laptop if I moved it to `parallel` and ran 32 competing node test processes. Even for a flaky test, that's unusually low. I typically don't see problems, even for flaky tests, until I get up to around four times that number. On a Raspberry Pi, of course, that sensitivity to resource constraints will manifest much sooner. This change checks the order of timers firing, rather than the duration before a timer is fired. This eliminates the sensitivity to resource constraints. The test can now be moved back to `parallel`. I am able to run many copies of the test simultaneously without seeing test failures. PR-URL: nodejs#10903 Fixes: nodejs#10073 Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com> Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Merged
italoacasas
pushed a commit
to italoacasas/node
that referenced
this pull request
Jan 30, 2017
Even after being moved to `sequential` in 1ce05ad, `test-regress-nodejsGH-897` still was occasionally flaky on Raspberry Pi devices on CI. The test is especially sensitive to resource constraints. It failed reliably on my laptop if I moved it to `parallel` and ran 32 competing node test processes. Even for a flaky test, that's unusually low. I typically don't see problems, even for flaky tests, until I get up to around four times that number. On a Raspberry Pi, of course, that sensitivity to resource constraints will manifest much sooner. This change checks the order of timers firing, rather than the duration before a timer is fired. This eliminates the sensitivity to resource constraints. The test can now be moved back to `parallel`. I am able to run many copies of the test simultaneously without seeing test failures. PR-URL: nodejs#10903 Fixes: nodejs#10073 Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com> Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
italoacasas
pushed a commit
to italoacasas/node
that referenced
this pull request
Jan 30, 2017
Even after being moved to `sequential` in 1ce05ad, `test-regress-nodejsGH-897` still was occasionally flaky on Raspberry Pi devices on CI. The test is especially sensitive to resource constraints. It failed reliably on my laptop if I moved it to `parallel` and ran 32 competing node test processes. Even for a flaky test, that's unusually low. I typically don't see problems, even for flaky tests, until I get up to around four times that number. On a Raspberry Pi, of course, that sensitivity to resource constraints will manifest much sooner. This change checks the order of timers firing, rather than the duration before a timer is fired. This eliminates the sensitivity to resource constraints. The test can now be moved back to `parallel`. I am able to run many copies of the test simultaneously without seeing test failures. PR-URL: nodejs#10903 Fixes: nodejs#10073 Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com> Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
MylesBorins
pushed a commit
that referenced
this pull request
Mar 8, 2017
Even after being moved to `sequential` in 1ce05ad, `test-regress-GH-897` still was occasionally flaky on Raspberry Pi devices on CI. The test is especially sensitive to resource constraints. It failed reliably on my laptop if I moved it to `parallel` and ran 32 competing node test processes. Even for a flaky test, that's unusually low. I typically don't see problems, even for flaky tests, until I get up to around four times that number. On a Raspberry Pi, of course, that sensitivity to resource constraints will manifest much sooner. This change checks the order of timers firing, rather than the duration before a timer is fired. This eliminates the sensitivity to resource constraints. The test can now be moved back to `parallel`. I am able to run many copies of the test simultaneously without seeing test failures. PR-URL: #10903 Fixes: #10073 Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com> Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
MylesBorins
pushed a commit
that referenced
this pull request
Mar 9, 2017
Even after being moved to `sequential` in 1ce05ad, `test-regress-GH-897` still was occasionally flaky on Raspberry Pi devices on CI. The test is especially sensitive to resource constraints. It failed reliably on my laptop if I moved it to `parallel` and ran 32 competing node test processes. Even for a flaky test, that's unusually low. I typically don't see problems, even for flaky tests, until I get up to around four times that number. On a Raspberry Pi, of course, that sensitivity to resource constraints will manifest much sooner. This change checks the order of timers firing, rather than the duration before a timer is fired. This eliminates the sensitivity to resource constraints. The test can now be moved back to `parallel`. I am able to run many copies of the test simultaneously without seeing test failures. PR-URL: #10903 Fixes: #10073 Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com> Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Merged
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Reviewers
6 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Even after being moved to sequential in
1ce05ad, test-regress-GH-897 still
was occasionally flaky on Raspberry Pi devices on CI.
The test is especially sensitive to resource constraints. It failed
reliably on my laptop if I moved it to parallel and ran 32 competing
node test processes. Even for a flaky test, that's unusually low. I
typically don't see problems, even for flaky tests, until I get up to
around four times that number.
On a Raspberry Pi, of course, that sensitivity to resource constraints
will manifest much sooner.
This change checks the order of timers firing, rather than the duration
before a timer is fired. This eliminates the sensitivity to resource
constraints. The test can now be moved back to parallel. I am able to
run many copies of the test simultaneously without seeing test failures.
Fixes: #10073
Checklist
Affected core subsystem(s)
test timers