5 Nov
2020
5 Nov
'20
1:30 p.m.
Hi Oliver
Review this "step by step" guide to see if you forgot something:
BR
NFS:
1.
chmod +x cephadm
2.
./cephadm bootstrap
-
Record dashboard user & password printed out at the end
3.
ADD OTHER HOSTS (assuming 3+ total after adding)
4.
./cephadm shell
5.
ceph orch apply osd --all-available-devices
6.
ceph fs volume create test 1
7.
ceph orch apply mds test 3
8.
ceph nfs cluster create cephfs testnfs
9.
ceph nfs cluster info testnfs
-
(verify hostname, ip and port are listed)
-
Record ip and port for later
10.
ceph nfs export create cephfs test testnfs /cephfs
11.
ceph auth ls
-
(check “client.testnfs1” keyring is present)
12.
ceph nfs export get testnfs /cephfs
-
(should have output)
13.
rados -p nfs-ganesha -N testnfs get export-1 - testnfs/cephnfs
-
(check that export was successfully created)
14.
ceph nfs export ls testnfs
-
(should show pseudo path “/cephfs”)
15.
Verify nfs export exists on dashboard
-
Login to dashboard with credentials from bootstrap
-
URL will be https://{host-ip}:8443/
-
Navigate to NFS page
-
Table should contain the export you just created
16.
Exit shell
-
Command should just be “exit”
17.
systemctl status nfs-server
-
If service is listed as inactive, run “systemctl start nfs-server”
-
Run “systemctl status nfs-server”. Should now be active
18.
sudo mount -t nfs -o port={nfs-port} {nfs-ip}:/cephfs /mnt
-
Port and ip should be from “ceph nfs cluster info testnfs” command
ran earlier
Ex:
mount -t nfs -o port=2049 10.8.128.94:/cephfs /mnt/cephfs/
Then give mount command -> to check if its mounted
#mount
Output:
10.8.128.94:/cephfs on /mnt/cephfs type nfs4
(rw,relatime,seclabel,vers=4.2,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=10.8.128.94,local_lock=none,addr=10.8.128.94)
--
Juan Miguel Olmo Martínez
Senior Software Engineer
Red Hat <https://www.redhat.com/>
jolmomar(a)redhat.com
<https://www.redhat.com/>