test-http-writable-true-after-close failure on centos #15505
Description
- Version: master
- Platform: centos
- Subsystem: http
I've seen a number of failures in runs today, including one that I started on master without any changes of:
https://ci.nodejs.org/job/node-test-commit-linux/12529/nodes=centos5-32/console
https://ci.nodejs.org/job/node-test-commit-linux/12525/nodes=centos5-64/console
not ok 751 parallel/test-http-writable-true-after-close
---
duration_ms: 0.281
severity: fail
stack: |-
Mismatched <anonymous> function calls. Expected exactly 1, actual 0.
at Object.exports.mustCall (/home/iojs/build/workspace/node-test-commit-linux/nodes/centos5-32/test/common/index.js:482:10)
at ClientRequest.common.mustCall (/home/iojs/build/workspace/node-test-commit-linux/nodes/centos5-32/test/parallel/test-http-writable-true-after-close.js:24:28)
at ClientRequest.<anonymous> (/home/iojs/build/workspace/node-test-commit-linux/nodes/centos5-32/test/common/index.js:514:15)
at Object.onceWrapper (events.js:316:30)
at emitOne (events.js:115:13)
at ClientRequest.emit (events.js:210:7)
at HTTPParser.parserOnIncomingClient [as onIncoming] (_http_client.js:565:21)
at HTTPParser.parserOnHeadersComplete (_http_common.js:116:23)
at Socket.socketOnData (_http_client.js:454:20)
...
So we may have landed something that broke that without noticing.
Activity
mhdawson commentedon Sep 20, 2017
I think this landed recently #15404
mhdawson commentedon Sep 20, 2017
@mcollina can you take a look to see if that change might be related ?
mhdawson commentedon Sep 20, 2017
Looks like the test failed in the CI run for that change as well.
mcollina commentedon Sep 20, 2017
Ouch, my bad. There are always some red for any number of reasons :/. I will have a look tomorrow, if you want you can revert now and I will merge again when it’s passing.
mcollina commentedon Sep 22, 2017
@mhdawson I'm stuck with that. I can't make that test pass on CentOS. Can someone else with more direct experience on the environment have a look? I do not know why it passes everywhere else and not there. The
'close'event on'OutgoingMessage'is not fired there in that condition.1 remaining item
Trott commentedon Sep 25, 2017
@mcollina On CentOS 5 (in CI at least), it is calling the
finishevent rather than thecloseevent. Not sure if that information is helpful or not...test: fix http-writable-true-after-close flakyness
test: fix http-writable-true-after-close flakyness
test: fix http-writable-true-after-close flakyness
test: fix http-writable-true-after-close flakyness
test: fix http-writable-true-after-close flakyness