|
2016-11-27
§
|
| 06:20 |
<ori@tin> |
Synchronized php-1.29.0-wmf.3/api.php: Bandaid: make API reqs fail fast if User-Agent ~= Parsoid and Host ~= eu.wikipedia.org (duration: 00m 50s) |
[production] |
| 05:36 |
<ori> |
Commented-out lived-hack from mw1290; if we see memory growth now, Parsoid would be strongly implicated. |
[production] |
| 05:33 |
<ori> |
With Parsoid requests hacked to fail fast, mw1290 is not showing the kind of aggressive growth in memory usage we're seeing on other API servers |
[production] |
| 05:30 |
<godog> |
roll restarting hhvm across api_cluster when hhvm uses more than 40% of memory |
[production] |
| 05:21 |
<ori> |
Live-hacked api.php on mw1290 to die if request user-agent contains 'Parsoid'; restarted HHVM. |
[production] |
| 05:17 |
<godog> |
roll restarting hhvm across api_cluster when hhvm uses more than 40% of memory |
[production] |
| 04:57 |
<godog> |
roll-restart hhvm on api_appcluster for on machines with hhvm leaking memory |
[production] |
| 03:22 |
<godog> |
roll-restart hhvm across api_appserver |
[production] |
| 02:41 |
<godog> |
dumping hhvm backtraces and roll-restart on affected api machines |
[production] |
| 02:00 |
<l10nupdate@tin> |
LocalisationUpdate failed: git pull of core failed |
[production] |
|
2016-11-25
§
|
| 20:09 |
<Krinkle> |
mwscript deleteEqualMessages.php --wiki angwiki (T45917) |
[production] |
| 17:15 |
<jynus> |
drop database vewikimedia (deleted wiki) from sanitarium and its slaves |
[production] |
| 14:22 |
<Reedy> |
delete oathauth row on wikitech for user Liuxinyu970226 per T144805 |
[production] |
| 14:16 |
<Reedy> |
delete oathauth row on wikitech for user Shoichi per T144805 |
[production] |
| 11:05 |
<ema> |
uploaded libvmod-{netmapper,tbf,vslp} to carbon main component (T150660) |
[production] |
| 10:20 |
<_joe_> |
upgrading HHVM across codfw |
[production] |
| 09:23 |
<_joe_> |
upgraded hhvm on the debug hosts |
[production] |
| 08:58 |
<_joe_> |
uploading hhvm_3.12.7+dfsg-1+wmf4 to apt |
[production] |
| 08:53 |
<volans> |
restarting zotero on sca1003, almost out of RAM, puppet failing |
[production] |
| 08:52 |
<elukey> |
restarting Yarn and HDFS masters on analytics100[12] (Hadoop cluster) to complete the openjdk update |
[production] |
| 07:51 |
<marostegui> |
Stopping replication db1052 for maintenance - T151607 |
[production] |
| 02:22 |
<l10nupdate@tin> |
ResourceLoader cache refresh completed at Fri Nov 25 02:22:40 UTC 2016 (duration 4m 20s) |
[production] |
| 02:18 |
<l10nupdate@tin> |
scap sync-l10n completed (1.29.0-wmf.3) (duration: 06m 48s) |
[production] |
|
2016-11-24
§
|
| 17:25 |
<_joe_> |
turned off additional workers for htmlcacheupdate on commonswiki as the queue has reduced to acceptable sizes (T151196) |
[production] |
| 15:03 |
<ema> |
uploaded varnish 4.1.3-1wm4 to carbon main component, replacing version 3.0.6plus-wm9 (T150660) |
[production] |
| 14:47 |
<ema> |
uploaded varnishkafka 1.0.12-1 to carbon main component, replacing version 1.0.7-1 (T150660) |
[production] |
| 13:31 |
<akosiaris> |
balance the load between thumbor1001 and thumbor1002 evenly |
[production] |
| 13:31 |
<akosiaris@puppetmaster1001> |
conftool action : set/weight=10; selector: thumbor1001.eqiad.wmnet (tags: ['dc=eqiad', 'cluster=thumbor', 'service=thumbor']) |
[production] |
| 13:20 |
<akosiaris@puppetmaster1001> |
conftool action : set/weight=5; selector: thumbor1001.eqiad.wmnet (tags: ['dc=eqiad', 'cluster=thumbor', 'service=thumbor']) |
[production] |
| 13:04 |
<akosiaris@puppetmaster1001> |
conftool action : set/weight=20; selector: thumbor1001.eqiad.wmnet (tags: ['dc=eqiad', 'cluster=thumbor', 'service=thumbor']) |
[production] |
| 12:54 |
<gilles> |
restarting thumbor on thumbor1001 |
[production] |
| 12:49 |
<akosiaris> |
lower thumbor1001 load by 50% to easy debugging |
[production] |
| 12:48 |
<gilles> |
restarting thumbor on thumbor1001 |
[production] |
| 12:48 |
<akosiaris@puppetmaster1001> |
conftool action : set/weight=5; selector: thumbor1001.eqiad.wmnet (tags: ['dc=eqiad', 'cluster=thumbor', 'service=thumbor']) |
[production] |
| 12:36 |
<elukey> |
launched preferred-replica-election to re-add kafka1022 among the Topic partition leader brokers of the Analytics Kafka cluster (all metrics looks good) |
[production] |
| 11:41 |
<hoo> |
Killed the Wikidata JSON dump creation on snapshot1007: Wont succeed before Monday, due to T151356 |
[production] |
| 10:13 |
<_joe_> |
running commonswiki htmlCacheUpdate jobs on terbium to catch up with the backlog, monitoring caches for vhtcpd queue overflows T151196 |
[production] |
| 09:38 |
<marostegui> |
Stopping replication db1052 (depooled) for maintenance - T150960 |
[production] |
| 08:59 |
<marostegui> |
Deploy alter table S5 - dewiki.revision on db1092 (depooled) - T148967 |
[production] |
| 08:15 |
<_joe_> |
uploaded calico-cni 1.5.1 to jessie-wikimedia |
[production] |
| 07:32 |
<marostegui> |
Stopping MySQL db2070 for maintenance - https://phabricator.wikimedia.org/T149553 |
[production] |
| 02:35 |
<l10nupdate@tin> |
ResourceLoader cache refresh completed at Thu Nov 24 02:35:10 UTC 2016 (duration 5m 15s) |
[production] |
| 02:29 |
<l10nupdate@tin> |
scap sync-l10n completed (1.29.0-wmf.3) (duration: 10m 39s) |
[production] |
| 00:28 |
<reedy@tin> |
Synchronized php-1.29.0-wmf.3/extensions/CentralAuth/maintenance/populateLocalAndGlobalIds.php: Some perf related improvements (duration: 00m 45s) |
[production] |
| 00:12 |
<demon@tin> |
Synchronized docroot/foundation/: rm more junk (duration: 00m 45s) |
[production] |