-
-
Notifications
You must be signed in to change notification settings - Fork 925
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
9.1.0.0 hangs on syswrite #3868
Comments
Ok, I started to analyze this and then realized I was looking at There are three places it could be hung: selecting for write, writing the bytes it needs, and subsequently flushing the IO. We need to see the JVM trace. Can you get a JVM stack trace from one of these hung processes? Should be Is there any way you get get us something reproducible? |
I'll get back to you with the JVM trace, I'm going to be calling JRuby.reference(t).native_thread.getStackTrace where Regarding a reproducible test case, I'm making about ~10,000 requests a day to this particular server, which I don't have access to, and this is only happening once every 3 days, so I'm afraid I have no idea of what a reproducible case would be |
@mohamedhafez Yes, that should work fine too, and we'll get just the thread we want. As long as you're able to access the environment we'll have an easier time investigating. You may also want to grab a heap dump (jmap tool, jvisualvm tool) when you next see a hanging thread so we have a memory snapshot to explore. cc @kares and @mkristian since this is OpenSSL related. It may be possible to downgrade the jruby-openssl gem to see if that fixes this issue. I do not have any theories at the moment. |
Is there a way to maybe record any relevant data just for the thread in question? Unfortunately the whole heap for the process is about 30 GB, and if I remember correctly it hangs or at least takes way too long to get a dump with visual vm... So just as an update, this hasn't happened since the last time I posted here, so I guess its a lot more uncommon than I thought... I'll report back with the JVM trace next time this happens, and if it doesn't happen for a while I'm fine closing the issue until I can provide more info if you'd prefer |
Haven't seen this happen since... huh. I'll close until/if it happens again, and I'll post a JVM stack trace if so |
9.1.0.0 has been hanging on
syswrite
once every few days for me while making an http request, which wasn't happening with 9.0.4.0 or earlier. I'm almost certain its due to some kind of strange/unexpected/incorrect behavior of the server I'm making the request of, as it doesn't happen for days and then all of a sudden I get a handful at the same time. Here's the stack trace:The text was updated successfully, but these errors were encountered: