1176803 - http://support.allizom.org/ - 503

Status: RESOLVED FIXED

(Infrastructure & Operations Graveyard :: WebOps: Community Platform, task)

Description

[Ryan C [:ryanc] (UTC-4)]

Getting a 503 with this. Was brought to my attention around 2200 PST by :stephend.

Comment 1

[Stephen Donner [:stephend] Not actively reading bugmail]

:mythmon - is this a SUMO issue, or should Web Ops take a look?

Comment 2

[Michael Cooper [:mythmon]]

I don't know of any SUMO problems that could be causing this. I don't see any requests in New Relic (which is normal for stage), and I don't see any errors in Sentry which I would normally see in case there are python errors. This sounds like an error in Apace, Zeus, modwsgi, something like that.

Comment 3

[C. Liang [:cyliang]]

support.allizom.org should be back up and working.

It looks like the wsgi process died at some point.[1]   I've done a graceful restart of Apache; if this problem crops up again, we should do more digging about in the logs.


[1] [Tue Jun 23 17:03:48 2015] [error] [client 208.181.190.99] (11)Resource temporarily unavailable: mod_wsgi (pid=11613): Unable to connect to WSGI daemon process 'kitsune-ssl' on '/var/run/wsgi.32377.15.1.sock'.

Comment 4

[Stephen Donner [:stephend] Not actively reading bugmail]

(In reply to C. Liang [:cyliang] from comment #3)
> support.allizom.org should be back up and working.
> 
> It looks like the wsgi process died at some point.[1]   I've done a graceful
> restart of Apache; if this problem crops up again, we should do more digging
> about in the logs.
> 
> 
> [1] [Tue Jun 23 17:03:48 2015] [error] [client 208.181.190.99] (11)Resource
> temporarily unavailable: mod_wsgi (pid=11613): Unable to connect to WSGI
> daemon process 'kitsune-ssl' on '/var/run/wsgi.32377.15.1.sock'.

Thanks; our automation (and therefore the site) seems to have been fine, ever since.

Marking as FIXED since you did a graceful Apache restart; :mythmon, if you want to do root-cause analysis, please feel free to reopen.