|
2016-08-18
ยง
|
| 17:19 |
<gehel> |
elasticsearch eqiad recovery throttling back to standard 20mb recovery |
[production] |
| 17:15 |
<godog> |
reinstall ms-be1027 after ssd replaced T140374 |
[production] |
| 17:12 |
<chasemp> |
nodepool restart with new settings for rate/env/ready for jessie from puppet |
[production] |
| 16:56 |
<mobrovac> |
mathoid deploying 75606c71 |
[production] |
| 16:16 |
<chasemp> |
restart nodepool with STATSD_HOST env variable for test |
[production] |
| 15:52 |
<gehel> |
reset postgresql replication on maps1003 (correction, not maps1001) |
[production] |
| 15:51 |
<gehel> |
reset postgresql replication on maps1001 |
[production] |
| 15:21 |
<aude@tin> |
Synchronized wmf-config/Wikibase.php: Bump cache epoch for Wikidata (duration: 00m 49s) |
[production] |
| 15:12 |
<aude@tin> |
Synchronized wmf-config/InitialiseSettings.php: Enable lazy loaded images on mobile web (duration: 01m 00s) |
[production] |
| 14:41 |
<paravoid> |
cr1-eqiad and cr2-eqiad: replace obsolete PIM BFD statements with new family inet/inet6 ones |
[production] |
| 14:38 |
<paravoid> |
cr1-eqiad: JunOS, SCB and linecard upgrade is over |
[production] |
| 14:34 |
<paravoid> |
cr1-eqiad: activating PyBal/LVS BGP sessions |
[production] |
| 14:32 |
<paravoid> |
cr1-eqiad: activating Transit4/6 BGP sessions |
[production] |
| 14:31 |
<paravoid> |
cr1-eqiad: activating Private-Peer4/6 BGP sessions |
[production] |
| 14:31 |
<paravoid> |
cr1-eqiad: reenabling Private-Peer/Transit interfaces |
[production] |
| 14:29 |
<mark> |
cr2-eqiad: setting ae3.1019 inet vrrp priority to default (from 50) |
[production] |
| 14:27 |
<paravoid> |
cr1-eqiad: removing deprioritization of all VRRP groups (priority=50) |
[production] |
| 14:24 |
<paravoid> |
cr1-eqiad: reenabling xe-5/0/3 (link to pfw) and the Fundraising BGP group |
[production] |
| 14:24 |
<paravoid> |
cr1-eqiad: reenabling xe-4/2/0 (link to cr1-codfw) and xe-4/2/2 (link to cr2-knams) |
[production] |
| 14:20 |
<paravoid> |
cr1-eqiad: reenabling all asw row D interfaces |
[production] |
| 14:19 |
<paravoid> |
cr1-eqiad: reenabling all asw row B/C interfaces |
[production] |
| 14:18 |
<gehel> |
increasing elasticsearch recovery throttling to 40mb to speed up eqiad recovery |
[production] |
| 14:18 |
<paravoid> |
cr1-eqiad: reenabling all asw row A interfaces |
[production] |
| 14:12 |
<paravoid> |
cr1-eqiad: activate chassis redundancy graceful-switchover |
[production] |
| 13:57 |
<paravoid> |
cr1-eqiad: cmjohnson1 replacing both SCBs with SCBE2s and adding a new linecard |
[production] |
| 13:54 |
<paravoid> |
cr1-eqiad: shutting down both routing-engines and powering off |
[production] |
| 13:38 |
<paravoid> |
cr1-eqiad: setting "chassis network-services enhanced-ip" and rebooting both REs |
[production] |
| 13:20 |
<paravoid> |
cr1-eqiad: upgrading re1 and rebooting |
[production] |
| 13:18 |
<paravoid> |
cr1-eqiad: toggling mastership between routing-engines (re1->re0) |
[production] |
| 13:00 |
<jynus> |
reloaded dbproxy1010's haproxy service to point to the original master |
[production] |
| 12:58 |
<paravoid> |
cr1-eqiad: upgrading re0 and rebooting |
[production] |
| 12:56 |
<paravoid> |
cr2-eqiad: remove VRRPv3 backwards compatibility (delete protocols vrrp checksum-without-pseudoheader) |
[production] |
| 12:53 |
<paravoid> |
cr1-eqiad: deactivate chassis redundancy graceful-switchover |
[production] |
| 12:52 |
<paravoid> |
cr1-eqiad: disabling all asw row A-C interfaces |
[production] |
| 12:43 |
<paravoid> |
cr1-eqiad: disabling all asw row D interfaces |
[production] |
| 12:40 |
<gehel@tin> |
Synchronized wmf-config/InitialiseSettings.php: switching search to codfw (duration: 00m 56s) |
[production] |
| 12:40 |
<gehel> |
switching search traffic to codfw |
[production] |
| 12:13 |
<paravoid> |
cr1-eqiad: reenabling all asw row A/B/C interfaces |
[production] |
| 12:05 |
<paravoid> |
cr1-eqiad: disabling all asw row B/C interfaces |
[production] |
| 12:04 |
<paravoid> |
cr1-eqiad: disabling all asw row A interfaces |
[production] |
| 12:01 |
<paravoid> |
cr1-eqiad: disabling Transit/Fundraising interfaces |
[production] |
| 11:56 |
<paravoid> |
cr1-eqiad: deactivating Fundraising BGP session |
[production] |
| 11:54 |
<paravoid> |
cr1-eqiad: deactivating PyBal/LVS backup static routes |
[production] |
| 11:47 |
<paravoid> |
cr1-eqiad: deactivating Transit4/6 and Private-Peer4/6 BGP sessions |
[production] |
| 11:31 |
<moritzm> |
uploaded new Linux package for jessie-wikimedia to carbon (now based on 4.4.18) |
[production] |
| 11:06 |
<paravoid> |
cr1-eqiad: deactivating BGP sessions with PyBal/LVS |
[production] |
| 11:04 |
<paravoid> |
cr1-eqiad: disabling xe-4/2/0 (link to cr1-codfw) and xe-4/2/2 (link to cr2-knams) |
[production] |
| 10:51 |
<paravoid> |
cr1-eqiad: deprioritizing all groups (priority=50) |
[production] |
| 10:20 |
<moritzm> |
upgrading remaining canary application servers to hhvm 3.12.7 |
[production] |
| 08:59 |
<moritzm> |
enabled scaling of huge SVGs on image scalers (T111815) |
[production] |