Degraded data redundancy: 7901/23703

[Palencia0289]

Since the first minute I created a CEPH cluster I have had this error
Degraded data redundancy: 7901/23703



root@nodo01:~# ceph osd tree
ID CLASS WEIGHT TYPE NAME STATUS REWEIGHT PRI-AFF
-1 1.59175 root default
-3 0.68217 host nodo01
0 hdd 0.45479 osd.0 up 1.00000 1.00000
2 hdd 0.22739 osd.2 up 1.00000 1.00000
-5 0.90958 host nodo02
1 hdd 0.45479 osd.1 up 1.00000 1.00000
3 hdd 0.45479 osd.3 up 1.00000 1.00000


root@nodo01:~# ceph -s
cluster:
id: 367caa66-be11-4c46-b0b0-4d4c7ea08b91
health: HEALTH_WARN
Degraded data redundancy: 7901/23703 objects degraded (33.333%), 129 pgs degraded, 129 pgs undersized

services:
mon: 2 daemons, quorum nodo02,nodo01 (age 46h)
mgr: nodo02(active, since 47h), standbys: nodo01
osd: 4 osds: 4 up (since 47h), 4 in (since 47h)

data:
pools: 2 pools, 129 pgs
objects: 7.90k objects, 31 GiB
usage: 63 GiB used, 1.5 TiB / 1.6 TiB avail
pgs: 7901/23703 objects degraded (33.333%)
129 active+undersized+degraded

io:
client: 10 KiB/s wr, 0 op/s rd, 1 op/s wr

root@nodo01:~#

 

[sb-jw]

CEPH needs at least 3 nodes and 3 mons / managers to work correctly. You obviously only have 2 nodes, hence the message.

 

[Palencia0289]

CEPH needs at least 3 nodes and 3 mons / managers to work correctly. You obviously only have 2 nodes, hence the message.

When you specify to work correctly, does that mean that my data is at risk with two nodes?

 

[VictorSTS]

When you specify to work correctly, does that mean that my data is at risk with two nodes?

Yes. And above that, the cluster won't work as you might expect:

- If one node or one OSD is down, no I/O will happen in your Ceph storage, as your min_size is 2 (assuming that you are using default settings).
- If data among the OSD's gets inconsistent for some reason, Ceph may not be able to tell which one of the OSD has the correct data.

For Ceph you need at least 3 nodes. If you only have two, take a look at storage replication instead [1].

[1] https://pve.proxmox.com/wiki/Storage_Replication