Hi, so - we're considering moving omap out to faster media than our rather slow spinning rust. There's been some discussion around this here: https://github.com/ceph/ceph/pull/6421
Since this hasn't landed in jewel, or the ceph-disk convenience bits - we're thinking of "other ways" of doing this. We're considering (ab)using partition UUIDs again, and symlinking it into place. This should at least avoid the "omap partition" weren't mounted, and tentatively have the OSD bail for that reason before accidentally doing something bad, and then amend the upstart unit to depend on the omap filesystem being mounted. Side-note, the question in https://github.com/ceph/ceph/pull/6421#issuecomment-152807595 regarding syncfs() and omap / objects being on separate devices, is this a legitimate concern ? If others have done something similar, I'd be happy to hear any experiences, great successes, failures or anything in between. Cheers, -- Kjetil Joergensen <kje...@medallia.com> SRE, Medallia Inc Phone: +1 (650) 739-6580
_______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
