test-cluster-primary-error flaky in test-asan job

[targos]

Example: https://github.com/nodejs/node/runs/3238780435?check_suite_focus=true

I think everybody knows about this one but I couldn't find an existing issue

[targos]

=== release test-cluster-primary-error ===
Path: parallel/test-cluster-primary-error
Error: Command: out/Release/node /home/runner/work/node/node/test/parallel/test-cluster-primary-error.js
--- TIMEOUT ---

===
=== 1 tests failed
===
make[1]: *** [Makefile:513: test-ci] Error 1

[lpinca]

Perhaps a longer timeout is sufficient?

[lpinca]

It's 5 minutes at the moment so I think it is already more than enough.

[tniessen]

FWIW, this is still one of the main reasons asan regularly fails, often taking hours for a full run (build + tests). Maybe we should temporarily increase the timeout to 10 minutes and see if that helps?

[mhdawson]

Another failure - #42459

[Trott]

Another failure - #42459

A link to the action results would be much more useful than a link to the PR, where the previous failing Action will be kind of hidden away. So here!: https://github.com/nodejs/node/runs/5684048217?check_suite_focus=true

[F3n67u]

Another failure: https://github.com/nodejs/node/runs/6958868902?check_suite_focus=true

[mmomtchev]

I have an environment where it fails every time, but only if it is run as part of the full suite, when running the test independently it passes. It fails along several other tests:
https://github.com/mmomtchev/libnode/runs/7118719199?check_suite_focus=true#step:7:11745
I will try to investigate, but if you want to give it a try:

git clone -b failing-unit-test https://github.com/mmomtchev/libnode.git
docker build -t yourname/libnode-ubuntu-bionic:latest .
docker run yourname/libnode-ubuntu-bionic:latest

[gengjiawen]

@theanarkh landed a fix on #44066.

[jasnell]

I'm assuming this is fixed now since #44066 landed? Please re-open if appears to still be flaky.