Project

General

Profile

Actions
Copy link

Bug #61582

open
« Previous | Next »

rbd unable to create snapshot: failed to allocate snapshot id: (95) Operation not supported

Added by Florius rk 11 months ago. Updated about 2 months ago.

Status:
New
Priority:
Normal
Assignee:
-
Category:
librbd
Target version:
-
% Done:

0%

Source:
other
Tags:
Backport:
Regression:
No
Severity:
2 - major
Reviewed:
Affected Versions:
v17.2.6
ceph-qa-suite:
rbd
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

Hi,

Running Proxmox 7.4-3 with Ceph version ceph version 17.2.6 (995dec2cdae920da21db2d455e55efbc339bde24) quincy (stable)
Kernel version Linux kvm-01 5.15.107-2-pve #1 SMP PVE 5.15.107-2 (2023-05-10T09:10Z) x86_64 GNU/Linux

I recently upgraded from 17.2.5 to 17.2.6.
Ever since when I try to create a RBD snapshot this fails with the following error:

# rbd snap create k8s_cephfs_data/csi-vol-8f127c1e-5dc8-4b9a-acc5-7f28eefd1ed0@test123
Creating snap: 10% complete...2023-06-04T00:59:34.873+0200 7f480695f700 -1 librbd::SnapshotCreateRequest: failed to allocate snapshot id: (95) Operation not supported
Creating snap: 10% complete...failed.
rbd: failed to create snapshot: (95) Operation not supported
root@kvm-01:~# rbd info k8s_cephfs_data/csi-vol-8f127c1e-5dc8-4b9a-acc5-7f28eefd1ed0
rbd image 'csi-vol-8f127c1e-5dc8-4b9a-acc5-7f28eefd1ed0':
        size 1 GiB in 256 objects
        order 22 (4 MiB objects)
        snapshot_count: 0
        id: 18ff5fe96f9e7d
        block_name_prefix: rbd_data.18ff5fe96f9e7d
        format: 2
        features: layering
        op_features:
        flags:
        create_timestamp: Mon May 15 20:32:39 2023
        access_timestamp: Mon May 15 20:32:39 2023
        modify_timestamp: Mon May 15 20:32:39 2023

This seems related to https://lists.ceph.io/hyperkitty/list/ceph-users@ceph.io/thread/TEHTDANDVT6P6BKCRHCABHFBM2PJFDJ5/#5RRA44GGQN5NQ2L35CCK3UPOLIO3R77S and https://forum.proxmox.com/threads/backup-vzsnap-fails-after-update-to-ceph-17-2-6.128265/

Actions
Copy link
 #1

Updated by Florius rk 11 months ago

This is done with an admin account:

client.admin
        key: <SNIP>
        caps: [mds] allow *
        caps: [mgr] allow *
        caps: [mon] allow *
        caps: [osd] allow *

Actions
Copy link
 #2

Updated by Florius rk 11 months ago

I created a new pool and making a snapshot on the new pool worked. Same type of image, and pool settings are the same:

root@kvm-01:~# rbd info test/test123
rbd image 'test123':
        size 1 GiB in 256 objects
        order 22 (4 MiB objects)
        snapshot_count: 0
        id: 58a4955b5797e2
        block_name_prefix: rbd_data.58a4955b5797e2
        format: 2
        features: layering
        op_features:
        flags:
        create_timestamp: Sun Jun  4 11:34:02 2023
        access_timestamp: Sun Jun  4 11:34:02 2023
        modify_timestamp: Sun Jun  4 11:34:02 2023
root@kvm-01:~# rbd info k8s_cephfs_data/csi-vol-785f53c1-1a96-4aa9-a9fb-b25654af9c83
rbd image 'csi-vol-785f53c1-1a96-4aa9-a9fb-b25654af9c83':
        size 700 GiB in 179200 objects
        order 22 (4 MiB objects)
        snapshot_count: 0
        id: dc65ed407e8fc1
        block_name_prefix: rbd_data.dc65ed407e8fc1
        format: 2
        features: layering
        op_features:
        flags:

create_timestamp: Sun Apr 16 00:26:42 2023
access_timestamp: Sun Apr 16 00:26:42 2023
modify_timestamp: Sun Apr 16 00:26:42 2023
Actions
Copy link
 #3

Updated by Florius rk 11 months ago

So I still had an old Ceph FS on that specific pool, k8s_cephfs_data.
I removed the FS and snapshots worked again.

This worked before <17.2.6. Is this a bug or intended behaviour? Please advice, thank you!

Actions
Copy link
 #4

Updated by Paul Kusters about 2 months ago

Issue still exists on Ceph version 18.2.1, for me removing the CephFS on the affected pool isn't an option.

Actions
Copy link

Also available in: Atom PDF