Intermittent segfault on SIGTERM #3355
-
Rails: 3.0.5p211 Hi, I'm currently investigating an issue with a dockerized rails app that when sent a SIGTERM (graceful shutdown) most of the time it throws a segfault - I have had it gracefully shutdown before, but more often than not it doesn't. This is causing 502s/504s in our application as our load balancers/cluster is expecting the requests to complete. I'm not sure if this is an application related problem, where we have a gem that isn't handling forks/threads properly - or if its a ruby/puma related problem. Any insight would be greatly appreciated! I've attached the dump:
|
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
It is a problem in Ruby, not Puma. It looks like it has been reported at https://bugs.ruby-lang.org/issues/19865 already (the bug report explains very well what's happening). Even looks like there's a suggested patch for it (ruby/ruby@master...nobu:ruby:signal-at-cleanup). |
Beta Was this translation helpful? Give feedback.
It is a problem in Ruby, not Puma. It looks like it has been reported at https://bugs.ruby-lang.org/issues/19865 already (the bug report explains very well what's happening). Even looks like there's a suggested patch for it (ruby/ruby@master...nobu:ruby:signal-at-cleanup).