osd: crash from ballpit0
#2 Updated by Colin McCabe almost 9 years ago
It seems like we are getting some out-of-order messages, just like in bug 666.
Here is the one that crashed us:
2010-12-25 16:25:46.621172 7fc0daef9910 osd0 3498 pg[1.ed( v 3496'15634 (3494'15630,3496'15634] n=1 ec=2 les=440 437/437/437) [2,0] r=1 luod=0'0 lcod 3494'15631 active] sub_op_modify trans 200.00009fa6/head v 3496'15633 (transaction) 87
The message came from OSD 2. Unfortunately, we don't have a log for that OSD any more.