Hello everyone, For the second time now, we have a warning on a Ceph cluster about a large omap object. This object is supposed to be in default.rgw.log, except after running listomapkeys on every object of that pool (`for i in `rados --cluster=ceph-par -p default.rgw.log ls`; do echo -n "$i:"; rados --cluster=ceph-par -p default.rgw.log listomapkeys $i |wc -l; done`), we get 0 for every object. As I understand it, objects in that pool are short-lived so the object which triggered the warning does not even exist anymore. Am I mistaken? The first time, we tried quite a lot of things, like triggering a deep scrub on the relevant pg, even looking into other pools (just in case) if we could find object with large omap objects (and we did not). The warning ended up going away on its own after a bit less than a week iirc. Is this a bug? Is there a way to clear this warning (if it is indeed about a non-existing object)? Thanks, _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx
 |