GTEST/UD: Increase UD EP timeout when running under valgrind - v1.17.x #9913
+1
−0
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Outdated suggestions cannot be applied.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.
This is double commit of #9880, into v1.17.x branch
Fix for RM#3918537
I managed to reproduce this issue on rock machines in 100% of the cases, but only when running this test under high CPU load. This CPU load I generate using dummy 64 processes (yes > /dev/null).
I checked ud_ep timeout logic, and it seems to work correctly. So the reasonable fix would be to increase UCX_UD_TIMEOUT (from 30s to 300s) when running under valgrind. With increased timeout the issue is not reproducible anymore, even with artificial CPU load