I debated upgrading to 1.6.3 as it was released right before we shipped
out our image. I opted not to as we had not had time to test it.
Had to say if that was a good choice or not now. FWIW, this is fixed in
1.6.3:
The tail end of this: https://bugzilla.lustre.org/show_bug.cgi?id=13436
<https://bugzilla.lustre.org/show_bug.cgi?id=13436>
Has:
Lustre 1.6.3, which contains a fix for this issue, has been released.
That said I still don't know if it is a critical error. I will try and
research that to see if we need to update, and what that will mean to
us.
--
Andrew
________________________________
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Lundgren,
Andrew
Sent: Monday, November 26, 2007 2:20 PM
To: [email protected]
Cc: Lynette, Troy; Bashir, Muhammad
Subject: [Lustre-discuss] ldlm_cli_enqueue errors.
I have been working with lustre for a short time now. I several
configured clusters and am seeing this error messages in the logs:
LustreError: 14232:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
LustreError: 14254:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
LustreError: 14276:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
LustreError: 14298:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
LustreError: 14320:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
LustreError: 14342:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
LustreError: 14364:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
LustreError: 14386:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
LustreError: 14408:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
LustreError: 14430:0:(mdc_locks.c:420:mdc_enqueue())
ldlm_cli_enqueue: -2
I have dug in the list archives a bit but not seen any real
answer to what this means.
I am using RH/CentOS 5 with lustre 1.6.2
Linux HOST 2.6.18-8.1.8.el5_lustre.1.6.2smp #1 SMP Tue Aug 28
10:56:40 MDT 2007 x86_64 x86_64 x86_64 GNU/Linux
Is there an explanation for these? Are they something we should
be concerned about?
Thank you.
--
Andrew
_______________________________________________
Lustre-discuss mailing list
[email protected]
https://mail.clusterfs.com/mailman/listinfo/lustre-discuss
