|
2016-10-18
§
|
| 15:06 |
<bblack> |
upgrading nginx on all remaining cache_misc (eqiad, esams) - T144523 |
[production] |
| 14:54 |
<bblack> |
upgrading nginx on all cache_misc @ codfw - T144523 |
[production] |
| 14:54 |
<chasemp> |
rsync tools from labstore1001 to labstore1004 |
[production] |
| 14:43 |
<bblack> |
upgrading nginx on all cache_misc @ ulsfo - T144523 |
[production] |
| 14:40 |
<marostegui> |
Shutting down es2015 for hardware maintenance - T147769 |
[production] |
| 14:21 |
<bblack> |
upgrading nginx on cp4001 (cache_misc ulsfo) as prod canary |
[production] |
| 14:18 |
<bblack> |
uploading nginx-1.11.4+wmf3 to carbon jessie-wikimedia - T144523 |
[production] |
| 13:58 |
<jynus> |
restarting and upgrading db2049 and es2019 to test new config |
[production] |
| 13:53 |
<jynus> |
applying new init.d script on all mariadb 10 servers |
[production] |
| 12:52 |
<elukey> |
mw1169 back in service after reimage (MW Jobrunner) |
[production] |
| 11:55 |
<elukey> |
removed /etc/mysql/conf.d/research-client.cnf from stat1002 (root:root perms, not supposed to be there but only on stat1003) |
[production] |
| 11:37 |
<elukey> |
reimaging mw1169 to Debian Jessie (MW Jobrunner) |
[production] |
| 10:40 |
<elukey> |
mw1168.eqiad.wmnet back in service after reimage (MW Jobrunner) |
[production] |
| 09:28 |
<elukey> |
reimaging mw1168 to Debian Jessie (MW Jobrunner) |
[production] |
| 09:25 |
<elukey> |
varnishkafka restarting in upload/misc/maps with new settings (https://gerrit.wikimedia.org/r/316306) |
[production] |
| 09:18 |
<gehel> |
upgrade nodejs to 4.6.0 on maps2* servers |
[production] |
| 08:56 |
<moritzm> |
reimaging tin to jessie |
[production] |
| 08:53 |
<marostegui> |
Deploying ALTER table on S4 commonswiki (db1064 — last host) - T147305 |
[production] |
| 08:42 |
<jynus> |
clone db1052 -> db1053, will perform maintenance (db restarts, reboots on both) at the same time |
[production] |
| 07:57 |
<marostegui@mira> |
Synchronized wmf-config/db-eqiad.php: Depool db1064 as it needs an ALTER table and pool db1068 temporarily to serve vslow and dump service - T147305 (duration: 02m 53s) |
[production] |
| 03:19 |
<mutante> |
restarted grrrit-wm |
[production] |
| 03:18 |
<mutante> |
gerrit has logs now in /var/log/gerrit/ |
[production] |
| 03:15 |
<mutante> |
restarting gerrit for logging config change |
[production] |
| 02:37 |
<l10nupdate@tin> |
ResourceLoader cache refresh completed at Tue Oct 18 02:37:01 UTC 2016 (duration 5m 49s) |
[production] |
| 02:31 |
<mwdeploy@tin> |
scap sync-l10n completed (1.28.0-wmf.22) (duration: 10m 20s) |
[production] |
| 00:48 |
<bblack> |
restarting API hhvms with >40% mem usage via salt every 10 minutes in a loop from here forward. screen session on neodymium, named api-hhvm-restarts |
[production] |
| 00:39 |
<mutante> |
restarted hhvm on mw1281 (was at 47.7% usage) |
[production] |
| 00:31 |
<bblack> |
restarting hhvm on API nodes where it's using >30% mem |
[production] |
| 00:22 |
<bblack> |
restarting hhvm on *API* nodes where it's using >50% mem |
[production] |
| 00:22 |
<bblack> |
restarting hhvm on nodes where it's using >50% mem |
[production] |
| 00:05 |
<mutante> |
restarted hhvm on mw1194,mw1197,mw1198 |
[production] |
|
2016-10-17
§
|
| 23:27 |
<Pchelolo> |
running import deletions script on restbase1007 |
[production] |
| 22:26 |
<mutante> |
restarted gerrit on cobalt |
[production] |
| 22:07 |
<Pchelolo> |
running restriction import script on restbase1007 |
[production] |
| 20:58 |
<mutante> |
tegmen - stopped duplicate icinga-wm (ircecho) |
[production] |
| 20:53 |
<mutante> |
maintenance servers, terbium and wasat, now have IPv6 connectivity |
[production] |
| 20:33 |
<bearND> |
deployed mobileapps 13fa4b4 |
[production] |
| 20:32 |
<Krenair> |
updated status.wm.o apache config on wikitech-static box to correctly serve static assets again (T148438) |
[production] |
| 20:30 |
<bearND> |
starting mobileapps deploy |
[production] |
| 19:31 |
<cwd> |
disabled all dedupe jobs besides "contacts" |
[production] |
| 18:38 |
<gehel> |
deploying latest gui and binaries for wdqs |
[production] |
| 18:35 |
<Jeff_Green> |
switch payments-listener back to eqiad |
[production] |
| 18:17 |
<_joe_> |
dumping core on mw1194 |
[production] |
| 17:32 |
<Jeff_Green> |
switch payments-listener to codfw |
[production] |
| 17:20 |
<_joe_> |
restarting lvs on lvs1003/1006 for the api change |
[production] |
| 16:42 |
<ottomata> |
restarting hadoop nodemanagers 1 at a time |
[production] |
| 16:18 |
<ori> |
Restarted HHVM on API cluster EQIAD |
[production] |
| 15:32 |
<ottomata> |
rebootting analytics1030 |
[production] |
| 15:13 |
<elukey> |
ran kafka preferred-replica-election to allow kafka1018 to be back as broker replica leader |
[production] |
| 14:38 |
<elukey> |
mw1167 back in service after reimage (MW Jobrunner) |
[production] |