Hello Ceph team. I've recently (3 weeks ago) upgraded a Ceph cluster from Nautilus to Pacific (16.2.7), and have encountered a strange issue. Scrubs, either deep or not, are scheduled and are shown in the cluster status, but there is no disk IO and they never finish. At the moment all of my PGs have fallen behind on scrubs which has me worried since data can be inconsistent and I'm not aware. This seems similar to: https://tracker.ceph.com/issues/54172 but no solution was reported, it has not begun to work in my case. and https://lists.ceph.io/hyperkitty/list/ceph-users@xxxxxxx/thread/A45BWXLWC2PKLGA5G7GXKCZDNHEOL2LL/#RHHJXQCBV2GGTGD7GAX5PACEY6TFSWRQ Because I also see that message in the logs, but the mentioned backport has not happened as far as I can tell. I'm assuming that this is not common, otherwise there would be more reports, is there something special about my cluster? Any workarounds to have scrubs working again? Thank you and best regards, Filipe _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx
 |