Fatal ConnectTimeoutError when running multiple tests concurrently

[zjmletang]

commit 9faca5e (HEAD, tag: v0.17.4)
I've observed frequent ConnectTimeoutError across some tasks during concurrent execution. The logs are as follows:
AutoHCK (5).log

[kostyanf14]

@zjmletang I have zero ideas. The "10.0.2.15" is internal IP in network namespace. Looks like Studio VM hangs on for some reason.

[zjmletang]

@kostyanf14 Got it. I'll keep an eye on the Studio VM during my next concurrent test runs. If the issue recurs, I'll try to gather more logs or diagnostic info to see if I can find a pattern. Thanks!

[kostyanf14]

@zjmletang Do you see this issue again? We started running the downstream WHQL testing process in parallel (21 AutoHCK run in parallel) and do not see this issue.

[zjmletang]

@kostyanf14 I will get back to you in a few days based on the latest commit, since our HLK machine has encountered some issues.