[Gluster-devel] 3.4.0beta2 crash in conservative merge?

Vijay Bellur vbellur at redhat.com
Sat May 25 16:49:14 UTC 2013


On 05/25/2013 09:30 AM, Emmanuel Dreyfus wrote:
> Here is a client crash on 3.4.0beta2:
>
> [2013-05-25 03:33:00.044913] I
> [afr-self-heal-entry.c:2253:afr_sh_entry_fix] 0-gfs34-replicate-1:
> /manu/netbsd/usr/src/tools/gcc/obj/build/build-i386-unknown-ne
> tbsdelf6.0/libiberty: Performing conservative merge
> [2013-05-25 03:34:25.624657] W
> [client-rpc-fops.c:1994:client3_3_setattr_cbk] 0-gfs34-client-0: remote
> operation failed: No such file or directory
> [2013-05-25 03:34:25.625071] W
> [client-rpc-fops.c:1994:client3_3_setattr_cbk] 0-gfs34-client-1: remote
> operation failed: No such file or directory
> [2013-05-25 03:34:25.626138] W
> [client-rpc-fops.c:1755:client3_3_xattrop_cbk] 0-gfs34-client-1: remote
> operation failed: Undefined error: 0. Path: (null) (--)
> [2013-05-25 03:34:25.626373] W
> [client-rpc-fops.c:1755:client3_3_xattrop_cbk] 0-gfs34-client-0: remote
> operation failed: Undefined error: 0. Path: (null) (--)

>From the core file:
> Program terminated with signal 11, Segmentation fault.
> #0  0xb9e11ffd in dht_migration_complete_check_task ()
>     from /usr/pkg/lib/glusterfs/3.4.0beta2/xlator/cluster/distribute.so
> (gdb) bt
> #0  0xb9e11ffd in dht_migration_complete_check_task ()
>     from /usr/pkg/lib/glusterfs/3.4.0beta2/xlator/cluster/distribute.so
> #1  0xbb7ba3a6 in synctask_wrap () from /usr/pkg/lib/libglusterfs.so.0
> #2  0xbb4c99b0 in swapcontext () from /usr/lib/libc.so.12
>
> I can work on that, but I would need to know how to reliabily reproduce
> it.


Seems to be happening when rebalance was running and there was active 
I/O from this client. Can you check which dereference leads to this SEGV?

Thanks,
Vijay





More information about the Gluster-devel mailing list