8
osd journal size = 1000
9
filestore xattr use omap = true
12
mon addr = 192.168.11.5:6789
23
auth supported = cephx
27
mon addr = 192.168.11.5:6789
32
- Put setup to /etc/ceph/ceph.conf
33
- Create configuration directories
34
mkdir -p /var/lib/ceph/osd/ceph-0
35
mkdir -p /var/lib/ceph/mon/ceph-a
36
mkdir -p /var/lib/ceph/mds/ceph-a
37
- Generate cluster setup
38
mkcephfs -a -c /etc/ceph/ceph.conf -k ceph.keyring
39
it will generate /var/lib/ceph
40
- Check for complains in logs, some folders may not exists
42
- Start ceph, 3 daemons should run ceph-mon, ceph-mds, ceph-osd
43
- Check for complains. Some directories still may miss. And it
44
really drove me made with authorization...
45
- Declare we don't have replication
46
ceph osd pool set rbd size 1
47
ceph osd pool set data size 1
48
ceph osd pool set metadata size 1
49
- restart ceph afterwards
50
- ceph health should say OK
51
- connect from client and check if ceph health says OK
54
rbd create vol1 --size 1048576 (in MB)
55
- on the client (this will be just a block device, no sharing with other clients)
56
rbd map vol1 --pool rbd --name client.admin
57
mkfs.ext4 -m0 /dev/rbd/rbd/vol1
58
mount /dev/rbd/rbd/vol1 /mnt/ssd
59
- The shared file system is ceph fs and shall be used somehow like that:
60
mount -t ceph 192.168.15.5:6789:/ /mnt/ssd
61
but there are another auth issuse, etc. I have stoped.
b'\\ No newline at end of file'