[04:40:53] !log stashbot Restarted logstash on stashbot-logstash to reattach irc client to freenode [04:40:56] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Stashbot/SAL, Master [10:34:50] 10Tool-Labs-tools-Other, 7Epic: Convert all Labs tools to use cdnjs for static libraries and fonts - https://phabricator.wikimedia.org/T103934#1835768 (10Ricordisamoa) [15:03:03] 10Tool-Labs-tools-Other, 7Epic: Convert all Labs tools to use cdnjs for static libraries and fonts - https://phabricator.wikimedia.org/T103934#1835980 (10Ricordisamoa) [15:47:20] (03PS3) 10Ricordisamoa: Initial commit [labs/tools/faces] - 10https://gerrit.wikimedia.org/r/192096 [15:48:51] (03CR) 10Ricordisamoa: "PS3 changes tools-static.wmflabs.org/static to tools-static.wmflabs.org/cdnjs" [labs/tools/faces] - 10https://gerrit.wikimedia.org/r/192096 (owner: 10Ricordisamoa) [15:51:07] (03PS4) 10Ricordisamoa: Initial commit [labs/tools/faces] - 10https://gerrit.wikimedia.org/r/192096 [15:52:08] (03CR) 10Ricordisamoa: "PS4 updates jQuery from 2.1.0 to 2.1.4" [labs/tools/faces] - 10https://gerrit.wikimedia.org/r/192096 (owner: 10Ricordisamoa) [15:53:51] (03PS5) 10Ricordisamoa: Initial commit [labs/tools/faces] - 10https://gerrit.wikimedia.org/r/192096 [15:55:00] (03CR) 10Ricordisamoa: "PS5 updates Bootstrap from 3.2.0 to 3.3.6" [labs/tools/faces] - 10https://gerrit.wikimedia.org/r/192096 (owner: 10Ricordisamoa) [16:33:58] yuvipanda: around ? [16:35:03] tonythomas: yuvi was up until late last night, and it's only 8:30 am on the west coast, so probably not yet :-) [16:35:30] tonythomas: but maybe someone else can help you? [16:35:46] I see that /srv is 100% full in my 20 gig instance, https://dpaste.de/uXDw/raw and the redis logs shows up https://dpaste.de/v06M/raw [16:35:52] valhallasw`cloud: yeah ! can you please take a look ? [16:36:10] all started when tinajohnson reported getting [1cc77ed9] /wiki/Special:Newsletter/1/announce JobQueueError from line 719 of /srv/vagrant/mediawiki/includes/jobqueue/JobQueueRedis.php: Redis server error: Could not insert 1 EchoNotificationDeleteJob job(s). [16:37:43] and is that 1.5 gig out of 20 gig is allocated for /srv is normal ? [16:44:12] valhallasw`cloud: any pointers on what we are looking at ? [16:46:59] tonythomas: the entire instance is 20GB, and I think most of it is in / [16:47:01] let me check [16:47:09] valhallasw`cloud: okey [16:47:38] valhallasw`cloud: df -h https://dpaste.de/oAyB/raw [16:48:15] tonythomas: huh, that's odd. / is 20GB and then /srv is another 1.5G [16:48:38] its mapped from somewhere else ? [16:49:26] I'm not entirely sure how the openstack mounts are configured [16:50:08] is 1.5GB expected for your redis instance? i.e. should we get more space for redis, or should we prune what is in /srv? [16:51:02] valhallasw`cloud: we have our wiki there, but should be fine [16:51:21] there is a redis folder in /srv/ [16:53:28] let me rephrase. Should /srv be bigger or should the contents of /srv be smaller? In the first case, you either have to store /srv on / by reconfiguring your mounts (but that only gives you a few GB extra), or you have to move to a larger instance [16:54:11] I wished we got more out of the 20 gig instance in /srv [16:54:17] in the second case, there are methods to clean up redis a bit, but unless it can be wiped altogether it's a bit more fiddly [16:54:39] you can unmount /srv, and then /srv will be on the same partition as / [16:55:09] root@newsletter-test:/srv/redis# du -h redis-db.rdb [16:55:09] 9.9M redis-db.rdb [16:55:16] right, so redis is not the issue [16:55:35] valhallasw`cloud: okey. Just unmounting /srv will do it ? let me try [16:55:51] tonythomas: you'll need to move the data to /, of course, and fix up fstab to prevent it from being auto-mounted [16:55:58] which might bring you into a fight with puppet. not sure [16:56:23] valhallasw`cloud: puppet usually dont give up that easily ( form my previous battle with exim configurations ) [16:56:25] :\ [16:56:56] tonythomas: I don't know whether puppet configures fstab or not [16:57:04] but if there is no other way, and /srv is allocated less than 10% of the total space -- then :\ [17:06:54] valhallasw`cloud: and one more thing. the redis-server is not getting restarted [17:06:57] just stuck [17:08:06] tonythomas: could be because /srv is full. Try killing it forcefully? [17:08:39] valhallasw`cloud: I did restart the instance though. Same issue again [17:10:10] this is after unmounting /srv? [17:10:48] nope. before that [17:11:13] .... [17:11:27] but then /srv is still full? [17:12:35] valhallasw`cloud: yeah. shameless me never unmounted /srv yet [17:12:54] let me do that now [17:13:14] umount: /srv: device is busy. [17:14:16] right, mediawiki/apache and redis are probably still using it, so those would have to be stopped [17:14:35] note that you have to copy /srv as well [17:15:12] (cp -R /srv /srv_new, then after unmounting mv /srv_new /srv) [17:15:43] uh, cp -R -p /srv /srv_new to preserve ownership and permissions [18:03:53] 10Tool-Labs-tools-Other, 7Epic: Convert all Labs tools to use cdnjs for static libraries and fonts - https://phabricator.wikimedia.org/T103934#1836084 (10Ricordisamoa) [19:35:18] !log cvn Setting up Cubbie in cvn-app5. [19:35:21] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Cvn/SAL, Master