Increase timeout for testQUICEcho40KiBMultistreamWithMetrics

[agnosticdev]

Increase the metric fetching timeout to provide more leeway when multiple tests are running at one time.

[ekinnear]

Do we have evidence that this would have worked in 5 seconds vs. fixing it by making it faster?

[agnosticdev]

Do we have evidence that this would have worked in 5 seconds vs. fixing it by making it faster?

This has been seen only in internal test systems and I am not able to reproduce this. What I am wanting to do is increase the timeout here to detect if there is actually a logical problem taking place or if we are running into a timeout that was too low when the system is running concurrent tests. For example, for 2 seconds, if the system is having an issue or running tests concurrently then I can see where this could randomly timeout un load. I have a hard time believing that the same would be true for 5 seconds. If we still see the same random failures at 5 seconds then it's time to rewrite this logic in our test harness.