Forums

Home / Forums

You need to log in to create posts and topics. Login · Register

[SOLVED] Cannot add OSD's after replacing node


tntbizhub
14 Posts
October 20, 2024, 7:24 pm

Hi all!

After exchanging node san-02 (for better hardware) using the procedure to replace node I was unable to add the new disks installed on this server. I go to to the GUI to add the drives it says "Adding" after a few seconds the "adding" disappears and the disk not added.

I tried to add it mannually using this command that i got from Ceph documentation.

ceph osd create

sudo mkdir /var/lib/ceph/osd/ceph-{number from ceph osd create}

sudo mkfs -t ext4 /dev/sdX

sudo mount -o user_xattr /dev/sdX /var/lib/ceph/osd/ceph-{number from ceph osd create}

ceph-osd -i {number from ceph osd create} --mkfs --mkkey --no-mon-config (If I don't use --no-mon-config gives error on authentication from the cluster)

ceph auth add osd.{number from ceph osd create} osd 'allow *' mon 'allow rwx' -i /var/lib/ceph/osd/ceph-{number from ceph osd create}/keyring

ceph osd crush add osd.{number from ceph osd create} 0.1 root=default host=san-2 {name of the node as host}


But I wanted to add it the normal way... thru the GUI... because this way there's no LVM etc....

Has anyone had the same issue?

Thanks in advance.

Last edited on October 20, 2024, 11:04 pm by tntbizhub · #1

tntbizhub
14 Posts
October 20, 2024, 7:33 pm

Errors:

Running command: /usr/bin/ceph-authtool --gen-print-key
Running command: /usr/bin/ceph --cluster ceph --name client.bootstrap-osd --keyring /var/lib/ceph/bootstrap-osd/ceph.keyring -i - osd new 8610d8f4-d40b-40cb-9f18-10152f8ce718
stderr: 2024-10-20T20:30:53.471+0100 7ff992d9d700 -1 monclient(hunting): handle_auth_bad_method server allowed_methods [2] but i only support [2]
stderr: 2024-10-20T20:30:56.471+0100 7ff991d9b700 -1 monclient(hunting): handle_auth_bad_method server allowed_methods [2] but i only support [2]
stderr: 2024-10-20T20:30:59.475+0100 7ff992d9d700 -1 monclient(hunting): handle_auth_bad_method server allowed_methods [2] but i only support [2]
stderr: 2024-10-20T20:30:59.475+0100 7ff991d9b700 -1 monclient(hunting): handle_auth_bad_method server allowed_methods [2] but i only support [2]
stderr: 2024-10-20T20:31:05.475+0100 7ff991d9b700 -1 monclient(hunting): handle_auth_bad_method server allowed_methods [2] but i only support [2]
stderr: 2024-10-20T20:31:05.475+0100 7ff992d9d700 -1 monclient(hunting): handle_auth_bad_method server allowed_methods [2] but i only support [2]
stderr: 2024-10-20T20:31:14.475+0100 7ff99259c700 -1 monclient(hunting): handle_auth_bad_method server allowed_methods [2] but i only support [2]
stderr: 2024-10-20T20:31:14.475+0100 7ff991d9b700 -1 monclient(hunting): handle_auth_bad_method server allowed_methods [2] but i only support [2]
stderr: [errno 13] RADOS permission denied (error connecting to the cluster)
--> RuntimeError: Unable to create a new OSD id

#2

tntbizhub
14 Posts
October 20, 2024, 11:06 pm

Executed

cp /var/lib/ceph/bootstrap-osd/ceph.keyring /etc/ceph/ceph.client.bootstrap-osd.keyring
ceph auth import -i /etc/ceph/ceph.client.bootstrap-osd.keyring

and the error went away!

Source:https://access.redhat.com/solutions/2885161

Last edited on October 21, 2024, 7:12 pm by tntbizhub · #3

admin
2,929 Posts
October 23, 2024, 10:36 am

Good you were able to solve this.

Note we were not able to reproduce this issue. If we replace a node we can add new OSDs without issues.

#4