rbd map fails

[elias.steiner]

Hi There

We try to use the ceph cluster for persisten storage in our local okd installation with Rook.

The Operator creates the block storage correct in the ceph pool. But pods and also local clients are not able to map the storage.

rbd ls --id=admin -m 192.168.1.1:6789,192.168.1.3:6789,192.168.1.2:6789 --keyfile=KEYKEYKEY== okd-encrypted
csi-vol-0816b1a4-ce52-46c2-9feb-159a0d459706
csi-vol-3d56a63e-d8bd-4523-b642-e0d41f0f3e41
csi-vol-47eb54af-9e44-45b7-be48-7e660c3cfd4e
csi-vol-f6428a38-0b82-4450-87c9-1a35da3ef4e1

rbd map --id admin -m 192.168.1.1:6789,192.168.1.3:6789,192.168.1.2:6789 --keyfile=KEYKEYKEY== okd-encrypted/csi-vol-0816b1a4-ce52-46c2-9feb-159a0d459706 --device-type krbd --options noudev
rbd: sysfs write failed
In some cases useful info is found in syslog - try "dmesg | tail".
rbd: map failed: (110) Connection timed out

Hade some one simular problems?

 

[DNM]

If someone else finds this with google - see if this is the culprit https://github.com/ceph/ceph-csi/issues/3113 --
If you are not using the multus network or some pod network please remove netNamespaceFilePath from your value.yaml (if you are using helm) or from ceph-csi-config configmap

Had similar experience with OKD and ceph-csi installed with Helm.