Hi Robin, W dniu pt., 9.04.2021 o 19:26 Robin H. Johnson <robb...@gentoo.org> napisaĆ(a):
> Maciej had said they were going to create a new thread, but I didn't see > one yet. > > I want to start by noting problem was much worse on 2.2.8 & 2.2.9, and > that 2.2.13 & 2.3.9 don't get entirely hung at 100% anymore: a big > thanks for that initial work in fixing the issue. > After Christopher patch using another function (that does not allocate memory) to dump LUA stack trace the situation is much better. It used 100% cpu only once and only one thread, however HAProxy didn't kill itself like in your situation. Since then the issue did not occur again, I have anything useful to share. :( I'm using 2.2.11 HAProxy but will upgrade next week to 2.3, I'll report if anything interesting pops out. > We have struggled to reproduce the problem in testing environments, it > only turns up at the biggest regions, and plotting occurances of the > issue over the time dimension suggest that it might have some partial > correlation w/ a weird workload input. > On our side the issue occured on a machine with unusual workload, hundred thousands tcp connections and very little http requests/responses. We also use lua. Kind regards, Maciej