cephadm Adding OSD wal device on a new
Hi, After adding a node to the cluster (3 nodes) with cephadm, how do I add OSDs with the same configuration on the other nodes ? The other nodes have 12 drives for data osd-block AND 2 drives for wal osd-wal. There are 6 LVs in each wal disk for the 12 data drives. I have added the ODS with ceph orch daemon add osd hostname:/dev/nvme0n1 How do I attach the wal devices to the OSDs? I have the WAL volumes created nvme3n1 259:5 0 349.3G 0 disk |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--dad0df4e--149a--4e80--b451--79f9b81838b8 253:12 0 58.2G 0 lvm |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--a5f2e93a--7bf0--4904--a233--3946b855c764 253:13 0 58.2G 0 lvm |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--cc949e1b--2560--4d38--bc27--558550881726 253:14 0 58.2G 0 lvm |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--8846f50e--7e92--4f66--a738--ce3a89650019 253:15 0 58.2G 0 lvm |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--6d646762--483a--40ca--8c51--ea54e0684a94 253:16 0 58.2G 0 lvm `-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--74e58163--de1d--4062--a658--5b0356d43a87 253:17 0 58.2G 0 lvm How do I attach the wal volumes to the OSDs osd.36 nvme0n1 259:1 0 5.8T 0 disk `-ceph--3df7c5c3--c2c0--4498--9e17--2af79e448abc-osd--block--804b50ea--d44c--4cad--9177--8d722f737df9 253:0 0 5.8T 0 lvm Osd.37 nvme1n1 259:3 0 5.8T 0 disk `-ceph--30858acc--c48b--4a08--bb98--4c9b59112c59-osd--block--0a3a198b--66ec--4ed9--94da--fb171e190e38 253:1 0 5.8T 0 lvm nvme3n1 Thank you, Anantha
Hi, first, I'd recommend to use drivegroups [1] to apply OSD specifications to entire hosts instead of manually adding an OSD daemon. If you run 'ceph orch daemon add osd hostname:/dev/nvme0n1' then the OSD is already fully deployed, meaning wal, db and data device are all on the same disk (which is totally fine for NVMEs or SSDs, depending on the use-case). In that case you would need to move it to your target device by using the ceph-bluestore-tool: $ ceph-bluestore-tool bluefs-bdev-new-wal --path osd path --dev-target new-device The question is, if you already use NVMEs as OSD devices, where's the point in separating wal to different NVMEs? That doesn't make much sense to me, but if I misunderstand your setup, please clarify. [1] https://docs.ceph.com/en/quincy/cephadm/services/osd/#advanced-osd-service-s... Zitat von "Adiga, Anantha" <anantha.adiga@intel.com>:
Hi,
After adding a node to the cluster (3 nodes) with cephadm, how do I add OSDs with the same configuration on the other nodes ? The other nodes have 12 drives for data osd-block AND 2 drives for wal osd-wal. There are 6 LVs in each wal disk for the 12 data drives. I have added the ODS with ceph orch daemon add osd hostname:/dev/nvme0n1
How do I attach the wal devices to the OSDs? I have the WAL volumes created nvme3n1 259:5 0 349.3G 0 disk |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--dad0df4e--149a--4e80--b451--79f9b81838b8 253:12 0 58.2G 0 lvm |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--a5f2e93a--7bf0--4904--a233--3946b855c764 253:13 0 58.2G 0 lvm |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--cc949e1b--2560--4d38--bc27--558550881726 253:14 0 58.2G 0 lvm |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--8846f50e--7e92--4f66--a738--ce3a89650019 253:15 0 58.2G 0 lvm |-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--6d646762--483a--40ca--8c51--ea54e0684a94 253:16 0 58.2G 0 lvm `-ceph--75d65cd1--91e4--4a8f--869b--e2a550f83104-osd--wal--74e58163--de1d--4062--a658--5b0356d43a87 253:17 0 58.2G 0 lvm
How do I attach the wal volumes to the OSDs
osd.36 nvme0n1 259:1 0 5.8T 0 disk `-ceph--3df7c5c3--c2c0--4498--9e17--2af79e448abc-osd--block--804b50ea--d44c--4cad--9177--8d722f737df9 253:0 0 5.8T 0 lvm Osd.37 nvme1n1 259:3 0 5.8T 0 disk `-ceph--30858acc--c48b--4a08--bb98--4c9b59112c59-osd--block--0a3a198b--66ec--4ed9--94da--fb171e190e38 253:1 0 5.8T 0 lvm nvme3n1
Thank you,
Anantha
_______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
participants (2)
-
Adiga, Anantha
-
Eugen Block