[ceph-users] 1 osd Segmentation fault in test cluster

Sage Weil sage at newdream.net
Thu Oct 5 19:07:03 PDT 2017


On Sat, 30 Sep 2017, Marc Roos wrote:
> Is this useful for someone?

Yes!

>  1: (()+0xa29511) [0x7f762e5b2511]
>  2: (()+0xf370) [0x7f762afa5370]
>  3: (BlueStore::TwoQCache::_trim(unsigned long, unsigned long)+0x2df) 
> [0x7f762e481a2f]
>  4: (BlueStore::Cache::trim(unsigned long, float, float, float)+0x1d1) 
> [0x7f762e4543e1]
>  5: (BlueStore::MempoolThread::entry()+0x14d) [0x7f762e45a71d]

See http://tracker.ceph.com/issues/21259

The latest luminous branch (which you can get from 
https://shaman.ceph.com/builds/ceph/luminous/) has some additional 
debugging on OSD shutdown that should help me figure out what is causing 
this.  If this is something you can reproduce on your cluster, please 
install the latest luminous and set 'osd debug shutdown = true' in the 
[osd] section of your config, and then ceph-post-file the log after a 
crash.

Thanks!
sage




More information about the ceph-users mailing list