seanlinsley has joined #rubygems-aws
mocara has joined #rubygems-aws
mocara has quit [Ping timeout: 250 seconds]
mocara has joined #rubygems-aws
mocara has quit [Ping timeout: 246 seconds]
mocara has joined #rubygems-aws
mocara has quit [Ping timeout: 260 seconds]
mocara has joined #rubygems-aws
mocara has quit [Ping timeout: 241 seconds]
mocara has joined #rubygems-aws
mocara has quit [Read error: No route to host]
mocara1 has joined #rubygems-aws
mocara1 has quit [Ping timeout: 240 seconds]
mocara has joined #rubygems-aws
mocara has quit [Read error: Operation timed out]
mocara has joined #rubygems-aws
mocara has quit [Ping timeout: 264 seconds]
vertis has joined #rubygems-aws
vertis has quit [Quit: vertis]
mocara has joined #rubygems-aws
mocara has quit [Ping timeout: 252 seconds]
mocara has joined #rubygems-aws
mocara has quit [Ping timeout: 264 seconds]
mocara has joined #rubygems-aws
mocara has quit [Quit: Leaving.]
almostwhitehat has joined #rubygems-aws
GitHub5 has joined #rubygems-aws
GitHub5 has left #rubygems-aws [#rubygems-aws]
<
GitHub5>
[rubygems-aws] dwradcliffe opened pull request #152: Cleanup and simplify roles (master...simplify-roles)
http://git.io/XmUNdw
seanlinsley has quit [Quit: seanlinsley]
almostwhitehat has quit [Read error: Connection reset by peer]
almostwhitehat has joined #rubygems-aws
seanlinsley has joined #rubygems-aws
mocara has joined #rubygems-aws
mocara has quit [Quit: Leaving.]
<
dwradcliffe>
samkottler: any reason we still have balancer01 running?
<
samkottler>
dwradcliffe: nope
<
dwradcliffe>
I'm going to terminate it
<
samkottler>
dwradcliffe: thanks!
<
dwradcliffe>
(found it by tracking down stray nginx_status pings hitting the app server)
GitHub120 has joined #rubygems-aws
GitHub120 has left #rubygems-aws [#rubygems-aws]
<
GitHub120>
[rubygems-aws] skottler closed pull request #152: Cleanup and simplify roles (master...simplify-roles)
http://git.io/XmUNdw
GitHub133 has joined #rubygems-aws
<
GitHub133>
rubygems-aws/master e918719 Sam Kottler: Merge pull request #152 from dwradcliffe/simplify-roles...
<
GitHub133>
rubygems-aws/master ba5511e David Radcliffe: Cleanup and simplify roles. This shouldn't cause any changes on the servers.
GitHub133 has left #rubygems-aws [#rubygems-aws]
GitHub61 has joined #rubygems-aws
<
GitHub61>
rubygems-aws/master 3b7662b Evan Phoenix: Merge branch 'master' of github.com:rubygems/rubygems-aws
<
GitHub61>
rubygems-aws/master ef99b3e Evan Phoenix: Upgrade stat-update to version that doesn't save per-day stats for now
GitHub61 has left #rubygems-aws [#rubygems-aws]
<
dwradcliffe>
I added an alert for low memory
almostwhitehat has quit [Ping timeout: 252 seconds]
almostwhitehat has joined #rubygems-aws
almostwhitehat has quit [Remote host closed the connection]
<
dwradcliffe>
samkottler: you available to hack on chef stuff tonight?
mocara has joined #rubygems-aws
mocara has quit [Quit: Leaving.]
mocara has joined #rubygems-aws
mocara has quit [Client Quit]
mocara has joined #rubygems-aws
mocara has quit [Client Quit]
vertis has joined #rubygems-aws
<
vertis>
dwradcliffe: you around
<
dwradcliffe>
vertis: hey
<
vertis>
did you guys setup a free memory alert yesterday?
<
dwradcliffe>
today
<
vertis>
I got a triggered alert on balancer02
<
vertis>
when you had that redis problem?
<
vertis>
yesterday for me
<
dwradcliffe>
oh right :)
<
dwradcliffe>
maybe we need to adjust the threshold?
<
vertis>
free memory is slowly declining
<
dwradcliffe>
although it's clearly ... yeah that
<
vertis>
but I can't see anything really eating it
<
vertis>
the biggest virtual memory holder is rsyslogd
<
vertis>
ps -e -o pid,vsz,comm= | sort -n -k 2
<
dwradcliffe>
restart it?
<
vertis>
I was playing restarting whoopsie (which we probably don't need at all)
<
vertis>
but I'm not certain that it's an unusual figure
<
vertis>
it's more likely the nginx memory is creeping up
<
dwradcliffe>
what is whoopsie?
<
vertis>
ubuntu crash reporter
<
dwradcliffe>
ah, probably not something we need
<
vertis>
I'll bounce rsyslogd and see
<
vertis>
if it changes memory usage
<
vertis>
I think we should adjust the threshold to maybe 1GB
<
vertis>
I can't see anything wrong
<
vertis>
I'll watch it a little
<
vertis>
out of curiosity, why are we running redis ourselves still?
<
vertis>
elasticache supports redis now
<
vertis>
regarding yesterday
<
dwradcliffe>
probably the same reason we run postgresql ourselves
<
vertis>
been thinking about that
<
vertis>
I can't see how we could move without an outage
<
vertis>
which is annoying
<
dwradcliffe>
I think we should think about heroku postgres
<
vertis>
albeit a brief one
<
vertis>
that ties us to regions
<
vertis>
unless I'm wrong
seanlinsley has quit [Quit: seanlinsley]