updated the gist with the timestamp of every occurence, if that helps narrow down a problem
rvedotrc has quit
rvedotrc joined the channel
ruaok
yeah, we've been seeing random 504s. they are not just limited to one site, which suggests that it is a gateway problem
bitmap
ah
I'll add some code to retry to request there anyway
pingupingu has quit
pingupingu joined the channel
pingupingu has quit
pingupingu joined the channel
UmkaDK has quit
UmkaDK joined the channel
ruaok
Gentlecat: does the sentry setup catch 504s for meb on sakura?
reosarevok is in Spain now, will be somewhat more active :)
reosarevok
Still no mobile internet, mind
reosarevok answered all the emails for now
reosarevok sighs at 5 days of subscription edit mails :D
Gentlecat
ruaok: of its in the app, it should
if*
ruaok
ok, then it idn
isn't sakura's fault.
drsaunders has quit
ZarkBit joined the channel
D4RK-PH0ENiX has quit
D4RK-PH0ENiX joined the channel
D4RK-PH0ENiX has quit
D4RK-PH0ENiX joined the channel
adhawkins
Morning. Just came in to see if you were interested in another 504 report, but it seems like you already know about them :)
ariscop has quit
For what it's worth, I had another at 05:16:02 downloading a replication packet.
ruaok
yeah, we know.
at least we no know that it is a gateway problem and not a app server problem.
and I'd hate to stick zas on it, because we should be focusing on moving to NewHost, but zas remains on vacation.
so, shitty service will remain.
ruaok desperately hopes that a time will come when he won't feel super frustrated about work
adhawkins
It's not so bad to be honest. I only notice it because occasionally I get a failure email from my slave server. It seems to recover itself the following hour.
zas
Replication related 504s are due to Sakura, they happen when it swaps heavily. I'm working on openresty container atm
ruaok
and why is sakura swapping?
75 uwsgi processes running.
gee, I wonder.
adhawkins: let me know if it happens again
zas: does graphana have a graph for examining memory usage for a given server?