when loading orangefactor i get: 502: Bad Gateway loading https://brasstacks.mozilla.com/api/count?startday=2016-01-12&endday=2016-01-19&tree=trunk
ed, nigel mentioned that this might be something for you
Yeah it does this every now and again, I'll give it a kick. The whole stack is pretty ancient, it's still running python 2.6 etc, and a lot of the configs need some TLC.
[firstname.lastname@example.org ~]# /etc/init.d/orangefactor stop; /etc/init.d/nginx reload; /etc/init.d/orangefactor start stopping orangefactor [FAILED] Reloading nginx: [ OK ] starting orangefactorspawn-fcgi: child spawned successfully: PID: 2925 [ OK ]
The old process did not exist, so I was not able to see what caused the excessive memory usage.
Probably using supervisor could help improving its availability?
I've also pip installed the various requests[security] extras to stop the SSL warnings (since it's running Python 2.6) a la: https://urllib3.readthedocs.org/en/latest/security.html#pyopenssl (In reply to Xidorn Quan [:xidorn] (UTC+10) from comment #6) > Probably using supervisor could help improving its availability? Yeah likely, though in previous instances the process had hung rather than dieing entirely iirc. I also didn't set the box up, and most of the existing config is undocumented, so I'm a little wary about changing too much without documenting first, which would take longer than it's probably worth. I'd much prefer to just decommission the whole box, which we'll do once we have OrangeFactor v2 using Treeherder's API directly.