Master repeatedly killing workers due to timeouts
list at mrtech.ru
Sun Dec 11 09:19:48 EST 2011
I have a simillar problem that started after update, 3 days ago we
moved from REE 1.8 / Rails 2 to Ruby 1.9.3 and Rails 3.1, also we
replaced memcached with redis so this is huge update and there are
many places where we can have an issue.
I also have checked my MySQL slow log and there are no requests which
take more than 5 seconds to complete.
My question is it possible to log the last request (URI) made to the
worker before killing it on timeout?
It's really hard to track this problem, usually unicorn start killing
workers when we got 1500-2000 rpms on 8 workers.
E, [2011-12-11T18:11:42.335656 #31736] ERROR -- : worker=1 PID:1632
timeout (61s > 60s), killing
E, [2011-12-11T18:11:42.364076 #31736] ERROR -- : reaped
#<Process::Status: pid 1632 SIGKILL (signal 9)> worker=1
I, [2011-12-11T18:11:42.364364 #31736] INFO -- : worker=1 spawning...
I, [2011-12-11T18:11:42.366721 #5869] INFO -- : worker=1 spawned pid=5869
I, [2011-12-11T18:11:42.366942 #5869] INFO -- : Refreshing Gem list
More information about the mongrel-unicorn