On Wed, 12 Jun 2024 19:02:30 GMT, Pavel Rappo <pra...@openjdk.org> wrote:
> Preliminary testing on some affected configurations showed that even 20 > seconds might not be enough. Initially, I suggested that it would be hard to > imagine, yet here we are. I haven't seen the jtreg launch command or the jtr files for these timeout failures, but in the past on systems that Matthias has run into similar issues, it was observed that those systems run with a very high `-concurrency` value (which implies too many test processes running at the same time) which could cause slowness. I don't know if that's what is a contributing factor here. > A good solution would be to disable expect timeout and, thus, fall back onto > jtreg timeout. I think that's a better solution. That removes the part where we have to guess what's the best/appropriate timeout value to use in the test. ------------- PR Comment: https://git.openjdk.org/jdk/pull/19627#issuecomment-2164662278
