[Gluster-devel] Latest unstable (1.4) branch checkout seems... unstable

Brent A Nelson brent at phys.ufl.edu
Wed Aug 6 17:37:47 UTC 2008


Unfortunately, I can't check that issue, as I'm still having an issue 
which makes 1.4 completely unusable, lately...

>From "cp -a /usr /beast/usr0":
2008-08-06 13:22:53 C [client-protocol.c:223:call_bail] ns0-0: bailing 
transport2008-08-06 13:22:53 E 
[client-protocol.c:4122:protocol_client_cleanup] ns0-0: forced unwinding 
frame type(1) op(34) reply=@0xb233d9f8
2008-08-06 13:22:53 E [client-protocol.c:4122:protocol_client_cleanup] 
ns0-0: forced unwinding frame type(2) op(5) reply=@0xb233d9f8
2008-08-06 13:22:53 E [afr.c:3007:afr_close_cbk] ns0: (path=/usr0/bin/free 
child=ns0-0) op_ret=-1 op_errno=77(File descriptor in bad state)
2008-08-06 13:22:53 E [socket.c:1193:socket_submit] ns0-0: transport not 
connected to submit (priv->connected = 255)
2008-08-06 13:22:53 E [afr.c:3328:afr_stat_cbk] ns0: (child=ns0-0) 
op_ret=-1 op_errno=107(Transport endpoint is not connected)
2008-08-06 13:22:53 E [socket.c:712:socket_connect_finish] ns0-0: 
connection failed (Connection refused)
2008-08-06 13:23:03 C [client-protocol.c:223:call_bail] ns0-1: bailing 
transport2008-08-06 13:23:03 E 
[client-protocol.c:4122:protocol_client_cleanup] ns0-1: forced unwinding 
frame type(1) op(17) reply=@0xb23408c8
2008-08-06 13:23:03 E [afr.c:3007:afr_close_cbk] ns0: (path=/usr0/bin/free 
child=ns0-1) op_ret=-1 op_errno=107(Transport endpoint is not connected)
2008-08-06 13:23:03 E [client-protocol.c:4122:protocol_client_cleanup] 
ns0-1: forced unwinding frame type(1) op(0) reply=@0xb23408c8
2008-08-06 13:23:03 E [afr.c:3328:afr_stat_cbk] ns0: (child=ns0-1) 
op_ret=-1 op_errno=107(Transport endpoint is not connected)
2008-08-06 13:23:03 E [unify.c:147:unify_buf_cbk] mirrors: child(ns0): 
path(/usr0/bin): Transport endpoint is not connected
2008-08-06 13:23:03 E [socket.c:1193:socket_submit] ns0-1: transport not 
connected to submit (priv->connected = 0)

Apparently, the ns server doesn't complain.

I've also seen it complain about extended attributes not being support, 
even though they are, at least when I tried disabling acl support.

I even tried with btrfs, which lasts longer before freaking out in a 
similar manner, although it has an additional issue:

2008-08-06 13:16:06 E [afr.c:4180:afr_readdir_cbk] ns0: (child=ns0-0) 
op_ret=-1
op_errno=0(Success)
2008-08-06 13:16:06 E [fuse-bridge.c:1947:fuse_readdir_cbk] 
glusterfs-fuse: 7: READDIR => -1 (Success)
2008-08-06 13:16:27 E [afr.c:4180:afr_readdir_cbk] ns0: (child=ns0-0) 
op_ret=-1
op_errno=0(Success)
2008-08-06 13:16:27 E [fuse-bridge.c:1947:fuse_readdir_cbk] 
glusterfs-fuse: 21079: READDIR => -1 (Success)

2008-08-06 13:16:06 E [posix.c:2974:posix_readdir] ns0: telldir failed: 
Success
2008-08-06 13:16:27 E [posix.c:2974:posix_readdir] ns0: telldir failed: 
Success

As a result, GlusterFS can't ls/du when on top of btrfs (although I'm 
guessing that will be an easy one to fix).

Thanks,

Brent

PS GlusterFS seems very fast on btrfs, at least with metadata-intensive 
writes.

On Wed, 6 Aug 2008, Amar S. Tumballi wrote:

> Hi Brent,
> Vikas fixed some bugs in xattr related path of posix. (in patch-272), can
> you check now?
>
> Regards,
> Amar
>
> 2008/7/30 Brent A Nelson <brent at phys.ufl.edu>
>
>> FYI, ls works okay in directories that only contain other directories.  If
>> the directory contains files, it complains.  After that, the namespace
>> glusterfsd processes seem to hang altogether and require a kill -9.
>>
>> Thanks,
>>
>> Brent
>>
>>
>> On Wed, 30 Jul 2008, Brent A Nelson wrote:
>>
>>  On Wed, 30 Jul 2008, Vikas Gorur wrote:
>>>
>>>  Brent,
>>>>
>>>> Thanks for pin-pointing the patch. I tried to reproduce this with an
>>>> AFR+Unify setup. However, I haven't been able to yet. How easy is it to
>>>> reproduce this? Which operations did you do before it screwed up?
>>>>
>>>>
>>> Right after mounting, I find that df works, but the very first ls -al is
>>> extremely slow, complains, and gives bad results.  So it's failing right
>>> away, in my case.
>>>
>>> Thanks,
>>>
>>> Brent
>>>
>>>
>>
>> _______________________________________________
>> Gluster-devel mailing list
>> Gluster-devel at nongnu.org
>> http://lists.nongnu.org/mailman/listinfo/gluster-devel
>>
>
>
>
> -- 
> Amar Tumballi
> Gluster/GlusterFS Hacker
> [bulde on #gluster/irc.gnu.org]
> http://www.zresearch.com - Commoditizing Super Storage!
>





More information about the Gluster-devel mailing list