I see some test flakes due to the timeout being too strict, updating to a larger value. Adding tail -n 1, it looks like there may be leftover state for other runs. We really only care about one of the CLOSE_WAIT entries.
5.5 KiB
5.5 KiB