Hi,
I am running Lustre 1.6.0.1 and there are quite a few error
messages like the one below appearing on /var/log/messages -
LustreError: 3683:0:(ldlm_lib.c:1363:target_send_reply_msg()) @@@
processing error (-13) req@0000010075e7f400 x3320402/t0
o34->1886b0e3-2c11-8576-52ef-f84a4a205955@NET_0x2000081ad7644_UUID:-1
lens 312/128 ref 0 fl Interpret:/0/0 rc -13/0
The file system is working fine and there are no network
issues like dropped frames. Can this error be ignored or is there a way to
fix this? The lru_size is the default value on all the clients. Thanks
very much.
Regards
Balagopal
On Jun 28, 2007 09:37 -0300, Balagopal Pillai wrote:> I am running Lustre 1.6.0.1 and there are quite a few error > messages like the one below appearing on /var/log/messages - > > LustreError: 3683:0:(ldlm_lib.c:1363:target_send_reply_msg()) @@@ > processing error (-13) req@0000010075e7f400 x3320402/t0 > o34->1886b0e3-2c11-8576-52ef-f84a4a205955@NET_0x2000081ad7644_UUID:-1 > lens 312/128 ref 0 fl Interpret:/0/0 rc -13/0-13 is -EACCESS. I''m not sure why this is being printed on the console. Are there any other messages? Cheers, Andreas -- Andreas Dilger Principal Software Engineer Cluster File Systems, Inc.
On Thu, 28 Jun 2007, Andreas Dilger wrote:
Hi Andreas,
It happened only one more time after that yesterday -
LustreError: 3701:0:(ldlm_lib.c:1363:target_send_reply_msg()) @@@
processing error (-13) req@000001007dc95600 x3416614/t0
o34->1886b0e3-2c11-8576-52ef-f84a4a205955@NET_0x2000081ad7644_UUID:-1 lens
312/128 ref 0 fl Interpret:/0/0 rc -13/0
Jun 28 15:26:05 lustre-3ware kernel: LustreError:
3701:0:(ldlm_lib.c:1363:target_send_reply_msg()) Skipped 266 previous
similar messages
There are no other messages other than a few evictions
of clients when they were rebooted. What is the EACCESS error in Lustre?
Thanks.
Regards
Balagopal
> On Jun 28, 2007 09:37 -0300, Balagopal Pillai wrote:
> > I am running Lustre 1.6.0.1 and there are quite a few
error
> > messages like the one below appearing on /var/log/messages -
> >
> > LustreError: 3683:0:(ldlm_lib.c:1363:target_send_reply_msg()) @@@
> > processing error (-13) req@0000010075e7f400 x3320402/t0
> >
o34->1886b0e3-2c11-8576-52ef-f84a4a205955@NET_0x2000081ad7644_UUID:-1
> > lens 312/128 ref 0 fl Interpret:/0/0 rc -13/0
>
> -13 is -EACCESS. I''m not sure why this is being printed on the
console.
> Are there any other messages?
>
> Cheers, Andreas
> --
> Andreas Dilger
> Principal Software Engineer
> Cluster File Systems, Inc.
>
-------- Original Message --------
Hi Andreas,
I forgot to mention that the Lustre volumes are exported via
nfs to multiple Solaris and Linux servers.
Does the error messages have something to do with the nfs re-export?
Thanks
Balagopal
Andreas Dilger wrote:> On Jun 28, 2007 09:37 -0300, Balagopal Pillai wrote:
>
>> I am running Lustre 1.6.0.1 and there are quite a few error
>> messages like the one below appearing on /var/log/messages -
>>
>> LustreError: 3683:0:(ldlm_lib.c:1363:target_send_reply_msg()) @@@
>> processing error (-13) req@0000010075e7f400 x3320402/t0
>>
o34->1886b0e3-2c11-8576-52ef-f84a4a205955@NET_0x2000081ad7644_UUID:-1
>> lens 312/128 ref 0 fl Interpret:/0/0 rc -13/0
>>
>
> -13 is -EACCESS. I''m not sure why this is being printed on the
console.
> Are there any other messages?
>
> Cheers, Andreas
> --
> Andreas Dilger
> Principal Software Engineer
> Cluster File Systems, Inc.
>
On Jun 29, 2007 06:52 -0300, Balagopal Pillai wrote:> It happened only one more time after that yesterday - > > LustreError: 3701:0:(ldlm_lib.c:1363:target_send_reply_msg()) @@@ > processing error (-13) req@000001007dc95600 x3416614/t0 > o34->1886b0e3-2c11-8576-52ef-f84a4a205955@NET_0x2000081ad7644_UUID:-1 lens > 312/128 ref 0 fl Interpret:/0/0 rc -13/0 > Jun 28 15:26:05 lustre-3ware kernel: LustreError: > 3701:0:(ldlm_lib.c:1363:target_send_reply_msg()) Skipped 266 previous > similar messages > > There are no other messages other than a few evictions > of clients when they were rebooted. What is the EACCESS error in Lustre?Just the usual - some user is unable to access a particular file. The only unusual thing is why this error is being printed. The error is from the "MDS_GETATTR_NAME" RPC, so it may well be related to NFS serving. Please file a bug with details. Cheers, Andreas -- Andreas Dilger Principal Software Engineer Cluster File Systems, Inc.