[Lustre-discuss] changed server handles on OST and EIOs

[Thomas Roth]

Hi all,

we are suffering from communication breakdowns between our clients
(1.6.6 and 1.6.7.2) and our MDT/OSTs (1.6.7.1), where in the log we see
numerous messages of the type
> Lustre: 2571:0:(import.c:736:ptlrpc_connect_interpret())
OST000e_UUID at 10.12.110.218@tcp changed server handle from
0xb97cb41513ad2c30 to 0xb97cb41513af5c68 but is still in recovery

In addition, there are errors
> LustreError: 14836:0:(file.c:117:ll_close_inode_openhandle()) inode
71142604 mdc close failed: rc = -116

> LustreError: 14836:0:(file.c:1010:ll_glimpse_size()) obd_enqueue
returned rc -5, returning -EIO
> LustreError: 14836:0:(llite_mmap.c:210:ll_tree_unlock()) couldn't
unlock -5
> LustreError: 2568:0:(socklnd.c:1656:ksocknal_destroy_conn())
Incomplete receive of slops from 12345-10.12.119.136 at tcp, ip
10.12.119.136:988, with error


On the OSTs, I see
> kernel: LustreError: 18128:0:(ldlm_lib.c:1643:target_send_reply_msg())
@@@ processing error (-107)  req at ffff8101ed0d3400 x17321/t0
o400-><?>@<?>:0/0 lens 128/0 e 0 to 0 dl 1250708929 ref 1 fl
Interpret:H/0/0 rc -107/0


Does that mean the MDT/OST file systems are damaged (rc -5), or is it
just bad connections as suggested by the other errors?

Any idea what's going on here?

TIA,
Thomas

-- 
--------------------------------------------------------------------
Thomas Roth
Department: Informationstechnologie
Location: SB3 1.262
Phone: +49-6159-71 1453  Fax: +49-6159-71 2986

GSI Helmholtzzentrum für Schwerionenforschung GmbH
Planckstraße 1
D-64291 Darmstadt
www.gsi.de

Gesellschaft mit beschränkter Haftung
Sitz der Gesellschaft: Darmstadt
Handelsregister: Amtsgericht Darmstadt, HRB 1528

Geschäftsführer: Professor Dr. Horst Stöcker

Vorsitzende des Aufsichtsrates: Dr. Beatrix Vierkorn-Rudolph,
Stellvertreter: Ministerialdirigent Dr. Rolf Bernhardt