Bug #12184
closedsame osds repeatedly marked down
0%
Description
Hi, on a cluster running 0.94.1 , both ssd osds of 1 host are repeatedly marked down:
2015-06-29 15:26:18.806731 7f8a22aaf700 0 -- 10.143.16.2:6802/1755 >> 10.143.16.1:6800/1685 pipe(0x3f892000 sd=24 :44531 s=2 pgs=1718 cs=1 l=0 c=0x3f86b9c0).fault with nothing to send, going to standby
2015-06-29 15:26:18.822995 7f8a22cb1700 0 -- 10.143.16.2:6802/1755 >> 10.143.16.1:6802/1781 pipe(0x3f889000 sd=22 :60056 s=2 pgs=1747 cs=1 l=0 c=0x3f868840).fault with nothing to send, going to standby
2015-06-29 15:26:23.455718 7f8a368dc700 0 monclient: hunting for new mon
2015-06-29 15:26:23.462725 7f8a368dc700 0 log_channel(cluster) log [WRN] : map e10824 wrongly marked me down
2015-06-29 15:31:25.367436 7f8a368dc700 0 monclient: hunting for new mon
2015-06-29 15:36:28.549766 7f8a368dc700 0 monclient: hunting for new mon
I see that on both osds.
I rebooted the affected host, but the problem remains.. after a few minutes they are marked down again.
I don't see any other problems on the host..
nothing special to see in dmesg too.
What could be the problem here?
Updated by Samuel Just almost 9 years ago
- Status changed from New to Rejected
Looks like a network problem.