On Thu, Nov 8, 2018 at 8:13 AM Yuri Weinstein <yweinste@xxxxxxxxxx> wrote: > > This release is to address regressions in this ticket > http://tracker.ceph.com/issues/36686 > > Distro runs are still in progress > > Details of this release summarized here: > http://tracker.ceph.com/issues/36711#note-2 > > rados - PASSED > rgw - Casey Matt approve? see known bugs > rbd - Jason approved > krbd - PASSED > fs - Patrick approve? > kcephfs - Patrick approve? > multimds - Patrick approve? > knfs - depreceted > ceph-deploy - Sage, Vasu approve? see bugs As per sage from past analysis it looks like disks are not cleaning up the partition and they are mostly bluestore tests that failed(there are some that have passed as well), we have been running *zap* before every osd create, probably zap is not doing its job + cm-ansible also does the osd partition cleanup. I will run this today in ovh and if it passes there it can confirm its a disk cleanup issue or its outside ceh-deploy. All faillures due to osd's not up, we have old tracker for this. http://qa-proxy.ceph.com/teuthology/yuriw-2018-11-06_17:15:26-ceph-deploy-luminous-distro-basic-mira/3230597/teuthology.log osd: 2 osds: 0 up, 0 in > ceph-disk PASSED > upgrade/client-upgrade-hammer (luminous) PASSED > upgrade/client-upgrade-jewel (luminous) PASSED > upgrade/luminous-p2p PASSED > upgrade/jewel-x (luminous) PASSED > upgrade/kraken-x (luminous) - Josh approved ?, EOL > powercycle - Neha approve? > ceph-ansible - Sebastian > https://github.com/ceph/ceph-ansible/issues/3306 needs fixing. > ceph-volume - PASSED Alfredo approved? > (please speak up if something is missing) > > Thx > YuriW >
