PGs stuck in created state

[simon falicon <simonfalicon@xxxxxxxxx>]

Hello Ceph Users,

I have an issue with my ceph cluster, after one serious fail in four SSD (electricaly dead) I have lost PGs (and replicats) and I have 14 Pgs stuck.

So for correct it I have try to force create this PGs (with same IDs) but now the Pgs stuck in creating state -_-" :

~# ceph -s
     health HEALTH_ERR
            14 pgs are stuck inactive for more than 300 seconds
....

ceph pg dump | grep creating

dumped all in format plain
9.3    0    0    0    0    0    0    0    0    creating    2019-02-25 09:32:12.333979    0'0    0:0[20,26]    20    [20,11]    20    0'0    2019-02-25 09:32:12.333979    0'0    2019-02-25 09:32:12.333979
3.9    0    0    0    0    0    0    0    0    creating    2019-02-25 09:32:11.295451    0'0    0:0[16,39]    16    [17,6]    17    0'0    2019-02-25 09:32:11.295451    0'0    2019-02-25 09:32:11.295451
...

I have try to create new PG dosent existe before and it work, but for this PG stuck in creating state.

In my monitor logs I have this message:

2019-02-25 11:02:46.904897 7f5a371ed700  0 mon.controller1@1(peon) e7 handle_command mon_command({"prefix": "pg force_create_pg", "pgid": "4.20e"} v 0) v1
2019-02-25 11:02:46.904938 7f5a371ed700  0 log_channel(audit) log [INF] : from='client.? 172.31.101.107:0/3101034432' entity='client.admin' cmd=[{"prefix": "pg force_create_pg", "pgid": "4.20e"}]: dispatch

When I check map I have:

~# ceph pg map 4.20e
osdmap e428069 pg 4.20e (4.20e) -> up [27,37,36] acting [13,17]

I have restart OSD 27,37,36,13 and 17 but no effect. (one by one)

I have see this issue http://tracker.ceph.com/issues/18298 but I run on ceph 10.2.11.

So could you help me please ?

Many thanks by advance,

Sfalicon.

_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com