Closed saitejar closed 6 years ago
ceph-mon.ceph
(and check if your host's DNS is set to kube-dns service ip)?syslog
and dmesg
?@rootfs
Yes, I can ping ceph-mon.ceph
In dmesg, I see.
[29119.038942] libceph: resolve 'ceph-mon.ceph' (ret=0): 10.211.55.195:0
[29119.038947] libceph: bad option at 'secret'
Looks like I am using a wrong secret. But I am unsure which secret to use with admin
user.
These are the secrets that ceph-helm deployment has created.
pvc-ceph-conf-combined-storageclass
ceph-bootstrap-mgr-keyring
ceph-client-admin-keyring
ceph-bootstrap-mds-keyring
ceph-bootstrap-osd-keyring
ceph-bootstrap-rgw-keyring
ceph-mon-keyring
pvc-ceph-client-key
ceph-keystone-user-rgw
how did you provision cephfs PV? it looks mount picked up secret file over secret
@rootfs, you are right. I was giving the key file name instead of keyfile. I am able to mount now. Thanks a lot for your help.
Request for Help. I am using ceph-helm, also running mds with it.
ceph -s
shows that everything is up and running.I am unable to mount ceph fs to a pod.
I am using this ceph-test-fs.yaml
I am not sure what keyring to use to be able to mount. and also I did try all the secrets generated by ceph-helm deployment.
Can anyone help me with mounting ceph fs?
Here is the error log:
MountVolume.SetUp failed for volume "cephfs" : CephFS: mount failed: mount failed: exit status 32 Mounting command: systemd-run Mounting arguments: --description=Kubernetes transient mount for /var/lib/kubelet/pods/7d82c944-e01f-11e7-bf03-001c42d61047/volumes/kubernetes.io~cephfs/cephfs --scope -- mount -t ceph -o name=admin,secretfile=/etc/ceph/admin.secret ceph-mon.ceph:6789:/ /var/lib/kubelet/pods/7d82c944-e01f-11e7-bf03-001c42d61047/volumes/kubernetes.io~cephfs/cephfs Output: Running scope as unit run-10734.scope. mount: wrong fs type, bad option, bad superblock on ceph-mon.ceph:6789:/, missing codepage or helper program, or other error In some cases useful info is found in syslog - try dmesg | tail or so.