[Fwd: Jan 18 08:51:33 4A:san1 unix: WARNING: oplock 0xa8000000afca6a80: timed out]

[James Peach]

2009/2/9 Herb Lewis <hlewis at panasas.com>:
> This might be better answered on the samba tech list. I didn't
> have much to do with the kernel oplock code.
>
> -------- Original Message --------
> Subject:        Jan 18 08:51:33 4A:san1 unix: WARNING: oplock
> 0xa8000000afca6a80: timed out
> Date:   Mon, 09 Feb 2009 16:13:41 -0600
> From:   Ron Short <short at sgi.com>
> To:     Herb Lewis <hlewis at panasas.com>
> References:     <4990A87C.3010700 at panasas.com>
>
>
>
> Herb,
>
> Not sure if you can respond, but since I believe you were at SGI
> sometime ago, Can you provide any guidance on the following using SGI's
> version of Samba:
>
> Jan 18 08:51:33 4A:san1 unix: WARNING: oplock 0xa8000000afca6a80: timed out
> Jan 18 08:54:51 4A:san1 unix: WARNING: oplock 0xa80000044f6a5e80: timed out
>
> Are these messages anything to be concerned about

Hi Ron,

These messages mean that an oplock break occurred which started the
break timer on the vnode. However, the timer expired before the client
released the oplock. This means that the kernel allowed the breaking
operation to complete, potentially violating the cache coherency the
oplock was supposed to ensure.

This is a 3-way dance between the kernel, Samba and the remote SMB
client. Somewhere, someone put a foot wrong.

> and/or is there a
> "good" way to keep them from being reported in the SYSLOG?

Not that I know of. They could indicate a problem, or possibly just a
network blip

>
> thanks
>
> --
> Ron Short                                       email: short at sgi.com
> Solutions Architect                             office: 651/683-5680
> SGI Global Professional Services                fax: 651/683-5288
>
>

-- 
James Peach | jorgar at gmail.com