Hello,
I have installed a Ceph Cluster (quincy) with 3 nodes. The problem I am
facing with since days is, that new hosts added into my cluster did not
show the disks I wanted to use for OSDs.
For example one of my nodes has 2 disks (SSD 500G).
|/dev/sda| is used for the OS (Debian 11)
The second disk |/dev/sdb| should be used by ceph for a OSD. But Ceph
does not recognize my 2nd disk on this node:
Calling ceph-volume directly on the node:
|$ sudo cephadm ceph-volume inventory |
results in an empty list - no disks recognized. But on the other hand,
the command 'lsblk' shows that the disk exists:
|$ lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT sda 8:16 1 476.9G 0
disk ├─sda1 8:17 1 32G 0 part [SWAP] ├─sda2 8:18 1 1G 0 part /boot
└─sda3 8:19 1 443.9G 0 part / sdb 8:0 1 476.9G 0 disk |
So it looks to me that something with my partitions or logical volumes
is totally wrong. Or something else?
I tried this with several different hosts. It seems that nodes with
NVME/SSD have no problem. But this can be a wrong observation by myself.
So I mean I have already added OSDs successful from my initial node.
My question is: how can I figure out what is wrong with the server that
it does not show its disk to ceph? The hardware is not totally
identically. So maybe there is something with the controller? But how
can we analyze this?
Thanks for any hints
===
Ralph
--
*Imixs Software Solutions GmbH*
*Web:* www.imixs.com <http://www.imixs.com> *Phone:* +49 (0)89-452136 16
*Timezone:* Europe/Berlin - CET/CEST
*Office:* Agnes-Pockels-Bogen 1, 80992 München
Registergericht: Amtsgericht Muenchen, HRB 136045
Geschaeftsführer: Gaby Heinle u. Ralph Soika
*Imixs* is an open source company, read more: www.imixs.org
<http://www.imixs.org>
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
