I got this yesterday when copying data to the rbd,
,----
| Apr 24 02:43:28 servername kernel: rbd: rbd1: write 40000 at 44893c10000
(10000)
| Apr 24 02:43:28 servername kernel:
| Apr 24 02:43:28 servername kernel: rbd: rbd1: result -28 xferred 40000
| Apr 24 02:43:28 servername kernel:
| Apr 24 02:43:28 servername kernel: end_request: I/O error, dev rbd1, sector
9198755968
| Apr 24 02:43:28 servername kernel: rbd: rbd1: write 10000 at 44893c00000 (0)
| Apr 24 02:43:28 servername kernel:
| Apr 24 02:43:28 servername kernel: rbd: rbd1: result -28 xferred 10000
| Apr 24 02:43:28 servername kernel:
| Apr 24 02:43:28 servername kernel: end_request: I/O error, dev rbd1, sector
9198755840
`----
and found some full warning in osd log
,----
| 2014-04-24 02:58:44.742167 7f09eb386700 0 log [WRN] : OSD near full (95%)
| 2014-04-24 02:59:14.154581 7f09eb386700 0 log [WRN] : OSD near full (95%)
| 2014-04-24 02:59:44.563262 7f09eb386700 0 log [WRN] : OSD near full (95%)
| 2014-04-24 03:00:14.272743 7f09eb386700 0 log [WRN] : OSD near full (95%)
`----
Is this nearly full osd has something to with previous kernel error? I
am using
tgt version 1.0.24
ceph version 0.72.2 (a913ded2ff138aefb8cb84d347d72164099cfd60)
kernel version 3.10.37-1.el6.elrepo.x86_64
BTW, What is the most stable kernel that support RBD for CentOS 6.4?
--
_________________________________________
/ Debian Hint #12: The grep-dctrl package \
| provides a number of useful scripts for |
| quickly searching various package |
| control files (such as the available |
\ packages file). /
-----------------------------------------
\
\
_____ _______
____==== ]OO|_n_n__][. | |
[________]_|__|________)< | |
oo oo 'oo OOOO-| oo\\_ ~~~|~~~
+--+--+--+--+--+--+--+--+--+--+--+--+--+
_______________________________________________
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
