test: Fix test-cluster-worker-exit.js for AIX #3666
Closed
Conversation
This file contains 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
test fails intermittently due to the assertion that the 'disconnect' event should come before the 'exit' event. This is caused be the non-deteministic behaviour of pollset_poll[1] on AIX (see deps/uv/src/unix/aix.c). This API makes no garauntee for the order in which file descriptors are returned. On linux epoll_wait[2] is used, which also does not make a garauntee on order of file descriptors returned. In the failing case we recieve our file descriptor with a callback of uv__signal_event (which causes JavaScript to receive the exit event) before our file descriptor with uv__stream_io as its callback (which in turn causes JavaScript receive the disconnect event). This change simply removes the assertion that the disconnect event happens before exit event and processes the test regardless of which event comes first. [1] https://www-01.ibm.com/support/knowledgecenter/ssw_aix_71/com.ibm.ai x.basetrf1/pollset.htm [2] http://linux.die.net/man/2/epoll_pwait
|
LGTM |
|
This fix is appropriate provided everybody agrees that there is no guarantee in respect how the events should occur. If that is not the case, a change to Node would be required to provide the expected behavior. @bnoordhuis can you chime in. |
|
Ah yes, good point. My assumption is that would be a |
mscdex
added
cluster
|
@nodejs/ctc |
|
From https://nodejs.org/api/cluster.html#cluster_cluster_workers "A worker is removed from cluster.workers after the worker has disconnected and exited. The order between these two events cannot be determined in advance." I believe this to be a faulty test by design. There should be no reason why we should assert that a 'disconnect' happens before an 'exit' event |
|
@trevnorris @mscdex @chrisdickinson ... any thoughts on this one? |
|
Makes sense. If the CI is happy, LGTM |
|
CI appears to be terminally unhappy these days. Hopefully it'll be kicked back into shape soon. Looking through the code, I agree with @iwuzhere's interpretation. I'll go ahead and land. We can revisit if necessary if CI issues pop up once the CI is stable again. |
jasnell
pushed a commit
that referenced
this pull request
Nov 14, 2015
test fails intermittently due to the assertion that the 'disconnect' event should come before the 'exit' event. This is caused be the non-deteministic behaviour of pollset_poll[1] on AIX (see deps/uv/src/unix/aix.c). This API makes no garauntee for the order in which file descriptors are returned. On linux epoll_wait[2] is used, which also does not make a garauntee on order of file descriptors returned. In the failing case we recieve our file descriptor with a callback of uv__signal_event (which causes JavaScript to receive the exit event) before our file descriptor with uv__stream_io as its callback (which in turn causes JavaScript receive the disconnect event). This change simply removes the assertion that the disconnect event happens before exit event and processes the test regardless of which event comes first. [1] https://www-01.ibm.com/support/knowledgecenter/ssw_aix_71/com.ibm.ai x.basetrf1/pollset.htm [2] http://linux.die.net/man/2/epoll_pwait PR-URL: #3666 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com>
|
Landed in 8b57b31 |
|
landed in v4.x-staging in 0fb40b4 |
MylesBorins
pushed a commit
that referenced
this pull request
Nov 17, 2015
test fails intermittently due to the assertion that the 'disconnect' event should come before the 'exit' event. This is caused be the non-deteministic behaviour of pollset_poll[1] on AIX (see deps/uv/src/unix/aix.c). This API makes no garauntee for the order in which file descriptors are returned. On linux epoll_wait[2] is used, which also does not make a garauntee on order of file descriptors returned. In the failing case we recieve our file descriptor with a callback of uv__signal_event (which causes JavaScript to receive the exit event) before our file descriptor with uv__stream_io as its callback (which in turn causes JavaScript receive the disconnect event). This change simply removes the assertion that the disconnect event happens before exit event and processes the test regardless of which event comes first. [1] https://www-01.ibm.com/support/knowledgecenter/ssw_aix_71/com.ibm.ai x.basetrf1/pollset.htm [2] http://linux.die.net/man/2/epoll_pwait PR-URL: #3666 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com>
Fishrock123
pushed a commit
that referenced
this pull request
Nov 17, 2015
test fails intermittently due to the assertion that the 'disconnect' event should come before the 'exit' event. This is caused be the non-deteministic behaviour of pollset_poll[1] on AIX (see deps/uv/src/unix/aix.c). This API makes no garauntee for the order in which file descriptors are returned. On linux epoll_wait[2] is used, which also does not make a garauntee on order of file descriptors returned. In the failing case we recieve our file descriptor with a callback of uv__signal_event (which causes JavaScript to receive the exit event) before our file descriptor with uv__stream_io as its callback (which in turn causes JavaScript receive the disconnect event). This change simply removes the assertion that the disconnect event happens before exit event and processes the test regardless of which event comes first. [1] https://www-01.ibm.com/support/knowledgecenter/ssw_aix_71/com.ibm.ai x.basetrf1/pollset.htm [2] http://linux.die.net/man/2/epoll_pwait PR-URL: #3666 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com>
rvagg
pushed a commit
that referenced
this pull request
Dec 4, 2015
test fails intermittently due to the assertion that the 'disconnect' event should come before the 'exit' event. This is caused be the non-deteministic behaviour of pollset_poll[1] on AIX (see deps/uv/src/unix/aix.c). This API makes no garauntee for the order in which file descriptors are returned. On linux epoll_wait[2] is used, which also does not make a garauntee on order of file descriptors returned. In the failing case we recieve our file descriptor with a callback of uv__signal_event (which causes JavaScript to receive the exit event) before our file descriptor with uv__stream_io as its callback (which in turn causes JavaScript receive the disconnect event). This change simply removes the assertion that the disconnect event happens before exit event and processes the test regardless of which event comes first. [1] https://www-01.ibm.com/support/knowledgecenter/ssw_aix_71/com.ibm.ai x.basetrf1/pollset.htm [2] http://linux.die.net/man/2/epoll_pwait PR-URL: #3666 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com>
Closed
jasnell
pushed a commit
that referenced
this pull request
Dec 17, 2015
test fails intermittently due to the assertion that the 'disconnect' event should come before the 'exit' event. This is caused be the non-deteministic behaviour of pollset_poll[1] on AIX (see deps/uv/src/unix/aix.c). This API makes no garauntee for the order in which file descriptors are returned. On linux epoll_wait[2] is used, which also does not make a garauntee on order of file descriptors returned. In the failing case we recieve our file descriptor with a callback of uv__signal_event (which causes JavaScript to receive the exit event) before our file descriptor with uv__stream_io as its callback (which in turn causes JavaScript receive the disconnect event). This change simply removes the assertion that the disconnect event happens before exit event and processes the test regardless of which event comes first. [1] https://www-01.ibm.com/support/knowledgecenter/ssw_aix_71/com.ibm.ai x.basetrf1/pollset.htm [2] http://linux.die.net/man/2/epoll_pwait PR-URL: #3666 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com>
jasnell
pushed a commit
that referenced
this pull request
Dec 23, 2015
test fails intermittently due to the assertion that the 'disconnect' event should come before the 'exit' event. This is caused be the non-deteministic behaviour of pollset_poll[1] on AIX (see deps/uv/src/unix/aix.c). This API makes no garauntee for the order in which file descriptors are returned. On linux epoll_wait[2] is used, which also does not make a garauntee on order of file descriptors returned. In the failing case we recieve our file descriptor with a callback of uv__signal_event (which causes JavaScript to receive the exit event) before our file descriptor with uv__stream_io as its callback (which in turn causes JavaScript receive the disconnect event). This change simply removes the assertion that the disconnect event happens before exit event and processes the test regardless of which event comes first. [1] https://www-01.ibm.com/support/knowledgecenter/ssw_aix_71/com.ibm.ai x.basetrf1/pollset.htm [2] http://linux.die.net/man/2/epoll_pwait PR-URL: #3666 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com>
jBarz
added a commit
to ibmruntimes/node
that referenced
this pull request
Feb 19, 2017
cherry-pick following from master
Author: Imran Iqbal <imrani@ca.ibm.com>
Date: Wed Nov 4 17:08:07 2015 -0500
test: Fix test-cluster-worker-exit.js for AIX
test fails intermittently due to the assertion that the 'disconnect'
event should come before the 'exit' event. This is caused be the
non-deteministic behaviour of pollset_poll[1] on AIX
(see deps/uv/src/unix/aix.c). This API makes no garauntee for the
order
in which file descriptors are returned. On linux epoll_wait[2] is
used,
which also does not make a garauntee on order of file descriptors
returned. In the failing case we recieve our file descriptor with a
callback of uv__signal_event (which causes JavaScript to receive the
exit event) before our file descriptor with uv__stream_io as its
callback (which in turn causes JavaScript receive the disconnect
event).
This change simply removes the assertion that the disconnect event
happens before exit event and processes the test regardless of which
event comes first.
[1]
https://www-01.ibm.com/support/knowledgecenter/ssw_aix_71/com.ibm.ai
x.basetrf1/pollset.htm
[2] http://linux.die.net/man/2/epoll_pwait
PR-URL: nodejs/node#3666
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
test fails intermittently due to the assertion that the 'disconnect'
event should come before the 'exit' event. This is caused be the
non-deteministic behaviour of pollset_poll[1] on AIX
(see deps/uv/src/unix/aix.c). This API makes no garauntee for the order
in which file descriptors are returned. On linux epoll_wait[2] is used,
which also does not make a garauntee on order of file descriptors
returned. In the failing case we recieve our file descriptor with a
callback of uv__signal_event (which causes JavaScript to receive the
exit event) before our file descriptor with uv__stream_io as its
callback (which in turn causes JavaScript receive the disconnect event).
This change simply removes the assertion that the disconnect event
happens before exit event and processes the test regardless of which
event comes first.
[1] https://www-01.ibm.com/support/knowledgecenter/ssw_aix_71/com.ibm.aix.basetrf1/pollset.htm
[2] http://linux.die.net/man/2/epoll_pwait