-
Notifications
You must be signed in to change notification settings - Fork 1.5k
Closed
Description
Hi, today I tried runnig puma on production server (MRI, but still wanted to check out the memory footprint), but I'm getting Errno::ENOTCONN after running puma for some time (in the end site is not responding at all)
2014-09-19 18:11:12: Read error: #<Errno::ENOTCONN: Socket is not connected - getpeername(2)>
/<path>/shared/bundle/ruby/2.1.0/gems/puma-2.9.1/lib/puma/server.rb:444:in `peeraddr'
/<path>/shared/bundle/ruby/2.1.0/gems/puma-2.9.1/lib/puma/server.rb:444:in `normalize_env'
/<path>/shared/bundle/ruby/2.1.0/gems/puma-2.9.1/lib/puma/server.rb:469:in `handle_request'
/<path>/shared/bundle/ruby/2.1.0/gems/puma-2.9.1/lib/puma/server.rb:361:in `process_client'
/<path>/shared/bundle/ruby/2.1.0/gems/puma-2.9.1/lib/puma/server.rb:254:in `block in run'
/<path>/shared/bundle/ruby/2.1.0/gems/puma-2.9.1/lib/puma/thread_pool.rb:92:in `call'
/<path>/shared/bundle/ruby/2.1.0/gems/puma-2.9.1/lib/puma/thread_pool.rb:92:in `block in spawn_thread'
Running on shared host (can't really check nginx logs):
ruby 2.1.2p95 (2014-05-08 revision 45877) [x86_64-freebsd9.2]
FreeBSD 9.2-STABLE FreeBSD 9.2-STABLE #3
rvm 1.25.30 (stable) by Wayne E. Seguin <[email protected]>, Michal Papis <[email protected]> [https://rvm.io/]
Rails 4.1.6
Puma config (+ running on UNIX socket):
threads 1, 8
workers 2
preload_app!
worker_timeout 60
After running it for whole day and bunch of restarts I finally had to revert back to unicorn (running fine on same server for over a year). Any ideas what might be causing this? 😢
Metadata
Metadata
Assignees
Labels
No labels