I was able to bring the osd's up by looking at my other OSD node which is the exact same hardware/disks and finding out which disks map. But I still cant bring up any of the start ceph-disk@dev-sd* services... When I first installed the cluster and got the OSD's up, I had to run the following:
# sgdisk -t 1:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdb # sgdisk -t 2:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdb # sgdisk -t 3:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdb # sgdisk -t 4:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdb # sgdisk -t 5:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdb # sgdisk -t 1:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdc # sgdisk -t 2:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdc # sgdisk -t 3:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdc # sgdisk -t 4:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdc # sgdisk -t 5:45b0969e-9b03-4f30-b4c6-b4b80ceff106 /dev/sdc Do i need to run that again? Cheers, Mike On Tue, Nov 29, 2016 at 4:13 PM, Sean Redmond <sean.redmo...@gmail.com> wrote: > Normally they mount based upon the gpt label, if it's not working you can > mount the disk under /mnt and then cat the file called whoami to find out > the osd number > > On 29 Nov 2016 23:56, "Mike Jacobacci" <mi...@flowjo.com> wrote: > >> OK I am in some trouble now and would love some help! After updating >> none of the OSDs on the node will come back up: >> >> ● ceph-disk@dev-sdb1.service >> loaded failed failed Ceph disk activation: /dev/sdb1 >> ● ceph-disk@dev-sdb2.service >> loaded failed failed Ceph disk activation: /dev/sdb2 >> ● ceph-disk@dev-sdb3.service >> loaded failed failed Ceph disk activation: /dev/sdb3 >> ● ceph-disk@dev-sdb4.service >> loaded failed failed Ceph disk activation: /dev/sdb4 >> ● ceph-disk@dev-sdb5.service >> loaded failed failed Ceph disk activation: /dev/sdb5 >> ● ceph-disk@dev-sdc1.service >> loaded failed failed Ceph disk activation: /dev/sdc1 >> ● ceph-disk@dev-sdc2.service >> loaded failed failed Ceph disk activation: /dev/sdc2 >> ● ceph-disk@dev-sdc3.service >> loaded failed failed Ceph disk activation: /dev/sdc3 >> ● ceph-disk@dev-sdc4.service >> loaded failed failed Ceph disk activation: /dev/sdc4 >> ● ceph-disk@dev-sdc5.service >> loaded failed failed Ceph disk activation: /dev/sdc5 >> ● ceph-disk@dev-sdd1.service >> loaded failed failed Ceph disk activation: /dev/sdd1 >> ● ceph-disk@dev-sde1.service >> loaded failed failed Ceph disk activation: /dev/sde1 >> ● ceph-disk@dev-sdf1.service >> loaded failed failed Ceph disk activation: /dev/sdf1 >> ● ceph-disk@dev-sdg1.service >> loaded failed failed Ceph disk activation: /dev/sdg1 >> ● ceph-disk@dev-sdh1.service >> loaded failed failed Ceph disk activation: /dev/sdh1 >> ● ceph-disk@dev-sdi1.service >> loaded failed failed Ceph disk activation: /dev/sdi1 >> ● ceph-disk@dev-sdj1.service >> loaded failed failed Ceph disk activation: /dev/sdj1 >> ● ceph-disk@dev-sdk1.service >> loaded failed failed Ceph disk activation: /dev/sdk1 >> ● ceph-disk@dev-sdl1.service >> loaded failed failed Ceph disk activation: /dev/sdl1 >> ● ceph-disk@dev-sdm1.service >> loaded failed failed Ceph disk activation: /dev/sdm1 >> ● ceph-osd@0.service >> loaded failed failed Ceph object storage daemon >> ● ceph-osd@1.service >> loaded failed failed Ceph object storage daemon >> ● ceph-osd@2.service >> loaded failed failed Ceph object storage daemon >> ● ceph-osd@3.service >> loaded failed failed Ceph object storage daemon >> ● ceph-osd@4.service >> loaded failed failed Ceph object storage daemon >> ● ceph-osd@5.service >> loaded failed failed Ceph object storage daemon >> ● ceph-osd@6.service >> loaded failed failed Ceph object storage daemon >> ● ceph-osd@7.service >> loaded failed failed Ceph object storage daemon >> ● ceph-osd@8.service >> loaded failed failed Ceph object storage daemon >> ● ceph-osd@9.service >> loaded failed failed Ceph object storage daemon >> >> I did some searching and saw that the issue is that the disks aren't >> mounting... My question is how can I mount them correctly again (note sdb >> and sdc are ssd for cache)? I am not sure which disk maps to ceph-osd@0 >> and so on. Also, can I add them to /etc/fstab to work around? >> >> Cheers, >> Mike >> >> On Tue, Nov 29, 2016 at 10:41 AM, Mike Jacobacci <mi...@flowjo.com> >> wrote: >> >>> Hello, >>> >>> I would like to install OS updates on the ceph cluster and activate a >>> second 10gb port on the OSD nodes, so I wanted to verify the correct steps >>> to perform maintenance on the cluster. We are only using rbd to back our >>> xenserver vm's at this point, and our cluster consists of 3 OSD nodes, 3 >>> Mon nodes and 1 admin node... So would this be the correct steps: >>> >>> 1. Shut down VM's? >>> 2. run "ceph osd set noout" on admin node >>> 3. install updates on each monitoring node and reboot one at a time. >>> 4. install updates on OSD nodes and activate second 10gb port, reboot >>> one OSD node at a time >>> 5. once all nodes back up, run "ceph osd unset noout" >>> 6. bring VM's back online >>> >>> Does this sound correct? >>> >>> >>> Cheers, >>> Mike >>> >>> >> >> _______________________________________________ >> ceph-users mailing list >> ceph-users@lists.ceph.com >> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >> >>
_______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
