Lustre discuss - Mar 2008 - strange lustre errors

Hi,

             On a few of the hpc cluster nodes, i am seeing a new lustre 
error that is pasted below. The volumes are working fine and there
is nothing on the oss and mds to report.


LustreError: 5080:0:(import.c:607:ptlrpc_connect_interpret()) 
data3-OST0000_UUID at 192.168.2.98@tcp changed handle from 
0xfe51139158c64fae to 0xfe511392a35878b3; copying, but this may 
foreshadow disaster
LustreError: 5080:0:(import.c:607:ptlrpc_connect_interpret()) 
data4-OST0000_UUID at 192.168.2.98@tcp changed handle from 
0xfe51139158c6502c to 0xfe511392a35878c1; copying, but this may 
foreshadow disaster
LustreError: 5080:0:(import.c:607:ptlrpc_connect_interpret()) 
scratch2-OST0003_UUID at 192.168.2.99@tcp changed handle from 
0x9ee58a75fddf2834 to 0x9ee58a761d190470; copying, but this may 
foreshadow disaster
LustreError: 5080:0:(import.c:607:ptlrpc_connect_interpret()) 
scratch1-OST0003_UUID at 192.168.2.99@tcp changed handle from 
0x9ee58a75fddf2754 to 0x9ee58a761d190462; copying, but this may 
foreshadow disaster

                Any help in interpreting these error messages is much 
appreciated. The two lustre oss/mds/mgs servers have been running fine 
with an uptime for over a month now after the ldiskfs patch is applied 
as mentioned here -
https://bugzilla.lustre.org/show_bug.cgi?id=13438

              The version of Lustre is 1.6.3.


Thanks
Balagopal

Hello!

On Mar 6, 2008, at 10:57 AM, Balagopal Pillai wrote:
>             On a few of the hpc cluster nodes, i am seeing a new  
> lustre
> error that is pasted below. The volumes are working fine and there
> is nothing on the oss and mds to report.
> LustreError: 5080:0:(import.c:607:ptlrpc_connect_interpret())
> data3-OST0000_UUID at 192.168.2.98@tcp changed handle from
> 0xfe51139158c64fae to 0xfe511392a35878b3; copying, but this may
> foreshadow disaster
This is some serious condition.
Please see bug 14775 (https://bugzilla.lustre.org/show_bug.cgi? 
id=14775) for
the patch and explanations..

Bye,
     Oleg

The bugzilla entry mentions that the patch was to go into 1.6.4 - which version
are you running?

The current stable if I am not mistaken is 1.6.4-2...

Are there any bugs in that which we should patch before we use it in production?

Michael

 -----Original Message-----
From: 	Balagopal Pillai [mailto:pillai at mathstat.dal.ca]
Sent:	Thursday, March 06, 2008 07:57 AM Pacific Standard Time
To:	lustre-discuss at lists.lustre.org
Subject:	[Lustre-discuss] strange lustre errors

Hi,

             On a few of the hpc cluster nodes, i am seeing a new lustre 
error that is pasted below. The volumes are working fine and there
is nothing on the oss and mds to report.


LustreError: 5080:0:(import.c:607:ptlrpc_connect_interpret()) 
data3-OST0000_UUID at 192.168.2.98@tcp changed handle from 
0xfe51139158c64fae to 0xfe511392a35878b3; copying, but this may 
foreshadow disaster
LustreError: 5080:0:(import.c:607:ptlrpc_connect_interpret()) 
data4-OST0000_UUID at 192.168.2.98@tcp changed handle from 
0xfe51139158c6502c to 0xfe511392a35878c1; copying, but this may 
foreshadow disaster
LustreError: 5080:0:(import.c:607:ptlrpc_connect_interpret()) 
scratch2-OST0003_UUID at 192.168.2.99@tcp changed handle from 
0x9ee58a75fddf2834 to 0x9ee58a761d190470; copying, but this may 
foreshadow disaster
LustreError: 5080:0:(import.c:607:ptlrpc_connect_interpret()) 
scratch1-OST0003_UUID at 192.168.2.99@tcp changed handle from 
0x9ee58a75fddf2754 to 0x9ee58a761d190462; copying, but this may 
foreshadow disaster

                Any help in interpreting these error messages is much 
appreciated. The two lustre oss/mds/mgs servers have been running fine 
with an uptime for over a month now after the ldiskfs patch is applied 
as mentioned here -
https://bugzilla.lustre.org/show_bug.cgi?id=13438

              The version of Lustre is 1.6.3.


Thanks
Balagopal
_______________________________________________
Lustre-discuss mailing list
Lustre-discuss at lists.lustre.org
http://lists.lustre.org/mailman/listinfo/lustre-discuss
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.lustre.org/pipermail/lustre-discuss/attachments/20080306/5d377b3e/attachment-0002.html