These are chat archives for fanout/pushpin

31st
May 2018
Jorge S. Cuesta
@jorgecuesta
May 31 2018 17:39
Hi guys, someone have some kind of news about this issue? fanout/pushpin#47638
someone else has this kind of issue? I’m working with pushpin with Docker under OSX
it happen after a few days of work and few container restarts
the main problem is it keep ‘restarting’ container
after that error I just could stop/remove container and make a new one
last time, I go inside container, remove mongrel2_7999.pid and it start normally
@jkarneges
Justin Karneges
@jkarneges
May 31 2018 18:59
Hi @jorgecuesta , I see the bigger issue now. it seems if mongrel2 exits uncleanly then it doesn't remove its pid file, which could prevent starting if another process uses the same PID afterwards
so I guess this unclean exit issue is not just cosmetic but should really be fixed. will look into this
Jorge S. Cuesta
@jorgecuesta
May 31 2018 21:06
exactly that, but I don’t found a way to delete that file after an unclean exit
my real fear is use pushpin on production with that kind of issue
if I could help in anything just give me a touch
Justin Karneges
@jkarneges
May 31 2018 22:34
I've updated pushpin's runner program to simply delete the mongrel2 pid file if it exists. I think this is the right thing for the runner to do, regardless of mongrel2's own behavior, since the runner is acting as the parent process manager and if the runner wasn't running then mongrel2 shouldn't have been running either
until that's released, I think you could do a similar thing in your container. just delete the mongrel2 pid file (if it exists) before starting pushpin