|
2012-08-22
§
|
| 23:54 |
<binasher> |
rebooting db35 for kernel upgrade |
[production] |
| 23:50 |
<binasher> |
new s5 master repl pos - MASTER_LOG_FILE='db45-bin.000187', MASTER_LOG_POS=807116168 |
[production] |
| 23:49 |
<asher> |
synchronized wmf-config/db.php 'setting s5 to writeable' |
[production] |
| 23:47 |
<asher> |
synchronized wmf-config/db.php 'setting s5 to read-only and switching master to db45, return db39 to s3' |
[production] |
| 23:47 |
<binasher> |
preparing to rotate the s5 master to db45 (last of these for now) |
[production] |
| 23:36 |
<mutante> |
check_job_queue check fixed after Tim synced spence. zhwiki has large queue: zhwiki (118614). Nagios did not talk because it was a transition from one critical to another |
[production] |
| 23:26 |
<Tim> |
added spence to the mediawiki-installation dsh group |
[production] |
| 23:25 |
<binasher> |
rebooting db39 for kernel upgrade |
[production] |
| 23:19 |
<binasher> |
new s3 master repl pos: MASTER_LOG_FILE='db34-bin.000864', MASTER_LOG_POS=466194762 |
[production] |
| 23:18 |
<asher> |
synchronized wmf-config/db.php 'setting s3 to writeable' |
[production] |
| 23:16 |
<asher> |
synchronized wmf-config/db.php 'setting s3 to read-only, switching master to db34' |
[production] |
| 23:15 |
<binasher> |
preparing to rotate the s3 master to db34 |
[production] |
| 23:14 |
<aaron> |
synchronized wmf-config/CommonSettings.php 'Restored math directory to its intended value.' |
[production] |
| 23:13 |
<aaron> |
synchronized wmf-config/InitialiseSettings.php 'Restored math directory to its intended value.' |
[production] |
| 23:06 |
<K4-713> |
updated production civicrm to [[rev:2013|r2013]] |
[production] |
| 22:42 |
<mutante> |
rebooting srv208,srv238 |
[production] |
| 22:35 |
<mutante> |
re-enabling puppet on srv208 and srv238, dist-upgrading while at it |
[production] |
| 22:27 |
<K4-713> |
updated production civicrm to [[rev:2011|r2011]] |
[production] |
| 22:10 |
<mutante> |
srv208,srv238 - puppet running but it's not. looks like puppet is disabled using puppet agent --disable, on purpose? |
[production] |
| 21:39 |
<catrope> |
synchronized php-1.20wmf10/resources 'jQuery revert' |
[production] |
| 21:03 |
<mutante> |
adding Bugzilla https monitor to watchmouse |
[production] |
| 20:53 |
<mutante> |
dist-upgrading kaulen, new kernel (not rebooted yet), fixed SSL cert path in puppet, restart puppet |
[production] |
| 20:37 |
<mutante> |
stopped puppet on kaulen. manually fixed path to SSL cert so Apache would start. Bugzilla back for now |
[production] |
| 20:32 |
<mutante> |
powercycling kaulen. OOM death. |
[production] |
| 20:29 |
<mutante> |
kaulen unresponsive, bugzilla affected |
[production] |
| 20:26 |
<aaron> |
synchronized php-1.20wmf9/extensions/SwiftCloudFiles/php-cloudfiles-wmf/cloudfiles.php 'deployed 8fea21b10eaba28773c30c6952fc792f08bf09a8' |
[production] |
| 20:24 |
<aaron> |
synchronized wmf-config/filebackend.php 'Re-enabled swift token caching.' |
[production] |
| 20:20 |
<aaron> |
synchronized php-1.20wmf10/extensions/SwiftCloudFiles/php-cloudfiles-wmf/cloudfiles.php 'deployed 8fea21b10eaba28773c30c6952fc792f08bf09a8' |
[production] |
| 19:17 |
<aaron> |
synchronized php-1.20wmf10/includes/Revision.php |
[production] |
| 19:14 |
<aaron> |
synchronized php-1.20wmf10/includes/Revision.php 'bit logic is fun' |
[production] |
| 19:11 |
<aaron> |
synchronized php-1.20wmf10/includes/Revision.php 'debug logging' |
[production] |
| 19:08 |
<mutante> |
reinstalling zirconium to use LVM |
[production] |
| 19:05 |
<aaron> |
synchronized php-1.20wmf10/includes/Revision.php 'debug logging' |
[production] |
| 18:55 |
<Restarted> |
ircecho on fenari, logmsgbot was broken --Roan |
[production] |
| 18:52 |
<RoanKattouw> |
1.20wmf10 deployed to all non-Wikipedia projects around 20:04 UTC |
[production] |
| 17:44 |
<mutante> |
installing package upgrades on bast1001 |
[production] |
| 17:19 |
<paravoid> |
stopped supervisord & puppet on silver/zhen per preilly's request |
[production] |
| 17:00 |
<apergos> |
oh, for people following along, shot the delete job. apparently even deletes is enough to push ms5 over the edge, yay |
[production] |
| 16:50 |
<apergos> |
removing thumbs on ms5 unused on any project, for june-july uploads. running in screen as root |
[production] |
| 16:37 |
<^demon> |
bugzilla went down, but came back up. magic? |
[production] |
| 16:35 |
<Jeff_Green> |
restarted ns1 b/c it fell down after authdns-update |
[production] |
| 16:33 |
<Jeff_Green> |
adding dns for pay-lvs100[12] |
[production] |
| 16:32 |
<paravoid> |
deploy new squid config that removes ms5 from a swift fallback for thumbs |
[production] |
| 16:06 |
<cmjohnson1> |
ms-be6 powering down for memory testing |
[production] |
| 15:32 |
<notpeter> |
rebooting srv193, upgrade to precise |
[production] |
| 11:38 |
<mark> |
Removed /homewmf NFS mount on kaulen |
[production] |
| 06:14 |
<binasher> |
new s1 repl position: MASTER_LOG_FILE='db63-bin.000049', MASTER_LOG_POS=646903534 |
[production] |
| 06:14 |
<asher> |
synchronized wmf-config/db.php 'returing db38 as an s1 slave' |
[production] |
| 06:02 |
<binasher> |
shutting down mysql on db38 and rebooting for kernel upgrade |
[production] |