[12:26:01] !log toolsbeta deploy latest buildservice (T336050) [12:26:04] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Toolsbeta/SAL [12:26:04] T336050: Buildpack pipeline: do not require a 'web' command to be present - https://phabricator.wikimedia.org/T336050 [12:26:16] !log toolsbeta deploy latest buildservice (T335865) [12:26:19] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Toolsbeta/SAL [12:26:19] T335865: PHP and Scala Heroku upstream buildpacks crash if ran with the Toolforge tool uid (instead of uid=1000) - https://phabricator.wikimedia.org/T335865 [12:28:40] !log tools deploy latest buildservice (T336050) [12:28:42] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [12:28:45] !log tools deploy latest buildservice (T335865) [12:28:47] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [15:11:38] !log codesearch temporarily taking down to free up disk space (T337263) [15:11:41] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Codesearch/SAL [15:11:41] T337263: full disk on codesearch8 - https://phabricator.wikimedia.org/T337263 [15:25:45] !log tools.poty-stuff Updated from de4b61d to 6e90969 [15:25:47] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.poty-stuff/SAL [20:38:41] I have mis-allocated 100G and need to move it from wikiweb/sdc on wikiapiary.wikiapiary to wikidb/sdb on wikiapaiary.wikiapiary. I thought of using horizion but it only lets you grow, not shrink. [20:38:53] opening phab ticket... [20:42:22] https://phabricator.wikimedia.org/T337434 [20:42:30] !help [20:42:30] If you don't get a response in 15-30 minutes, please create a phabricator task -- https://phabricator.wikimedia.org/maniphest/task/edit/form/1/?projects=wmcs-kanban [20:42:50] well, I just did that, so now I wait :) [20:57:14] hexmode: is there data that you need to move, or just empty volumes that need to be resized? Either way I think there are only a few folks who can help. andrewbogott is one. d.caro and d.hinus also, but they are in EU timezones if this is urgent. [20:57:57] not urgent [20:58:25] I can move the data to off the smaller disk [20:58:41] let me do that and update the task [20:59:20] "Amount to increase: move 100G from wikiweb/sdc to wikiweb/sdb" does not make a lot of sense to me yet. Is there another way to phrase the help you need? [21:02:04] bd808: I don't want an increase. I just want to move 100G from sdc to sdb on wikiapiary [21:02:46] ok. so you need help resizing existing cinder volumes? [21:02:53] yes [21:05:00] updated task with your clearer wording [21:07:45] even better :) [21:19:13] hexmode: I'm looking at the upstream docs, and the only resize they describe is to make things larger. But... it looks like you are not actually using the "wikidb" volume yet? [21:20:22] I am.... it is almost full. I just have it unmounted [21:20:32] ah, ok [21:20:45] my magic ideas wouldn't work then. ;) [21:21:02] I saw your note about snapshotting. I thought of that, but don't have the space atm [21:21:39] so the next thing for me was to move the data, resize, and then move it back [21:23:30] *nod* If you stash the ~20G on the wikiweb volume somewhere then you can delete that volume, grow the wikidb volume, and then make a new volume with the remaining quota to be the new wikiweb. [21:26:59] hexmode: could you rsync/scp the data from the wikiweb volume over to the dbbackup volume on the scripts instance temporarily? Or maybe even better mount that dbbackup volume on the wikiapiary instance temporarily as the stash? [21:27:39] * bd808 is looking for options that keep him from messing up cinder things but unblock hexmode [21:45:17] Sorry, went to take a shower for an award thing for my daughter... [21:45:34] but I will be syncing to another cloud instance, yeah [21:49:11] hexmode: is the issue just that you need temporary quota in order to shuffle things around? [21:53:40] andrewbogott: I think that could unblock him, yes. Basically he has 200G and 300G volumes with data in them and really wants 100G and 400G volumes instead. The 200G volume is only ~20G used. The 300G volume is almost full. [21:54:08] So just having someplace to stash that 20G would fix everything... [21:54:12] * andrewbogott checks quotas [21:56:47] That would be one way to handle it, but things are already in motion. I was hoping for a "simple" resize2fs, but was scared of these openstack volumes [21:58:05] Well, I raised the quota, just comment on ticket when you're done rearranging. [21:58:23] Generally it's easy to make volumes bigger and hard/impossible to make them smaller. [21:59:20] k