On Thu, 11 Jul 2013, Mandell Degerness wrote: > Is there any command (in shell or python API), that can tell me if > ceph is still creating pgs other than actually attempting a > modification of the pg_num or pgp_num of a pool? I would like to > minimize the number of errors I get and not keep trying the commands > until success, if possible. > > Right now, I run the command "ceph osd pool set <pool> pg_num <val>" > and immediately attempt he "ceph osd pool set <pool> pgp_num <val>" > command, repeating the latter until success happens or I get an error > other than "still creating pgs" > > It would be nice to run the first command, wait for the cluster to be > ready again, then run the second. You can get this now from 'ceph status --format=json' but, looking at it now, the json needs to be restructured; we'll do that now so that dumpling will have something more sane. For now, just look for 'creating' in the 'pgmap' key. I'll fix it to print all that info in a structured way. > As a bonus, if there is a reasonable way to do this (expand pgs) > entirely in python without running any shell commands, that would be > awesome. The REST API endpoint just got mergd this week and can do everything that the CLI can do. It will be in dumpling as well. sage _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |