|
2009-04-28
§
|
| 10:19 |
<Tim> |
re-added srv57 to mediawiki-installation, was rogue and causing "unknown cluster" errors |
[production] |
| 07:59 |
<tstarling> |
synchronized php-1.5/db.php 'set the new cluster22 to be the sole ES write destination' |
[production] |
| 07:57 |
<Tim> |
pdns on bayle is broken, stuck in futex, restarting |
[production] |
| 07:52 |
<tstarling> |
synchronized php-1.5/db.php |
[production] |
| 07:49 |
<tstarling> |
synchronized php-1.5/db.php 'introducing cluster22 (ms3/ms2)' |
[production] |
| 07:43 |
<Tim> |
adding tables called blobs_cluster22 to ms3, for new current text cluster |
[production] |
| 07:30 |
<Tim> |
fixed /etc/mysql/debian.cnf on ms3 so that logrotate flush logs can work |
[production] |
| 02:09 |
<andrew> |
synchronized php-1.5/CommonSettings.php 'Rolling out tor changes' |
[production] |
| 02:07 |
<andrew> |
synchronized php-1.5/InitialiseSettings.php 'Rolling out tor changes, and ipblock-exempt on all wikis' |
[production] |
| 01:48 |
<Andrew> |
Updating configuration to cchange tor settings. |
[production] |
|
2009-04-27
§
|
| 23:42 |
<tstarling> |
synchronized php-1.5/db.php 'gave the current ES masters some read load' |
[production] |
| 23:05 |
<Tim> |
increased connection limit on temp-es* from 100 to 500 |
[production] |
| 18:31 |
<Rob> |
srv138, srv139, & srv145 reinstalled and online. |
[production] |
| 18:24 |
<brion> |
stopped apache and umounted amane from srv184 (ES slave). load is way overloaded for some reason on this box |
[production] |
| 18:24 |
<Rob> |
removed amane from mounts on srv184 |
[production] |
| 18:01 |
<Rob> |
srv145 reinstalling |
[production] |
| 17:58 |
<Rob> |
some quirky stuff going on from various memcached hosts being reinstalled and such. Issues seem to be resolved now. |
[production] |
| 17:56 |
<robh> |
synchronized php-1.5/mc-pmtpa.php 'removing reinstalling servers' |
[production] |
| 17:54 |
<robh> |
synchronized php-1.5/mc-pmtpa.php 'removing reinstalling servers' |
[production] |
| 17:43 |
<Rob> |
srv129 back online |
[production] |
| 17:43 |
<Rob> |
reinstalling srv138 and srv139 |
[production] |
| 17:24 |
<Rob> |
srv126 up and online |
[production] |
| 17:11 |
<Rob> |
srv126 and srv129 being reinstalled. |
[production] |
| 17:09 |
<Rob> |
srv86 and srv87 up and online |
[production] |
| 16:49 |
<Rob> |
srv86 and srv87 upgrading to ubuntu |
[production] |
| 16:42 |
<Rob> |
srv107 online |
[production] |
| 16:38 |
<robh> |
synchronized php-1.5/mc-pmtpa.php 'Removing srv120-srv123 for other testing' |
[production] |
| 16:35 |
<robh> |
synchronized php-1.5/mc-pmtpa.php 'removing srv156' |
[production] |
| 16:22 |
<Rob> |
srv120-srv123 reinstalled, NOT online. Base OS, nothing else, passed on to mark for his testing. (Puppet I assume.) |
[production] |
| 15:48 |
<Rob> |
srv120-123 going down for reinstallation |
[production] |
| 15:45 |
<Rob> |
srv108 and srv109 up and online |
[production] |
| 15:06 |
<Rob> |
srv108 and srv109 are in mid-install for ubuntu |
[production] |
| 15:06 |
<Rob> |
srv107 wont restart for some reason, adding to tasks to troubleshoot. |
[production] |
| 15:04 |
<Rob> |
srv105 and srv106 back up and online |
[production] |
| 14:56 |
<Rob> |
srv107-srv109 goin down |
[production] |
| 14:54 |
<Rob> |
srv104 back online |
[production] |
| 14:48 |
<Rob> |
srv102 and srv103 back up and online |
[production] |
| 14:43 |
<Rob> |
srv102-106 reinstalling. |
[production] |
| 14:29 |
<Rob> |
srv53 has a bad fan, shutting down until its replaced. |
[production] |
| 14:20 |
<Rob> |
srv102-srv109 being upgraded to ubuntu. |
[production] |
| 11:43 |
<andrew> |
synchronized php-1.5/InitialiseSettings.php 'Updated $wgSitename for ukwikimedia in accordance with IRC request from Michael Peel, a board member' |
[production] |
| 02:20 |
<Tim> |
srv53 down, took it out of memcached rotation. Updating the memcached spare list. |
[production] |
| 02:20 |
<tstarling> |
synchronized php-1.5/mc-pmtpa.php |
[production] |
| 02:12 |
<Tim> |
fixed rc1 slaves, broken by expire_logs_days on ms3 |
[production] |
| 01:59 |
<Tim> |
Shut down srv217 for maintenance. Similar timer interrupt issue observed as before: select() syscalls running indefinitely despite a short timeout specified. |
[production] |
| 01:53 |
<tstarling> |
synchronized php-1.5/db.php |
[production] |
| 01:52 |
<Tim> |
repooled ms3 rc1 instance |
[production] |
| 01:49 |
<Tim> |
reset slave on db21, was running out of disk space due to relay logs |
[production] |
| 01:42 |
<Tim> |
fixed nagios for srv99, still had its apache check command set to my CGI security vulnerability demonstration, permanently saved in retention.dat despite config changes |
[production] |
| 01:17 |
<Tim> |
enabled apport on srv99, to see if I can track down the nagios flapping |
[production] |