[pnfs] Fwd: October Bakeathon: Draft 13 and NFSv4.1 OPEN clientid and other seqid/clientid questions.

[Lisa Week]

Hi Andy,

Responses inline.

Thanks,
Lisa

On Sep 18, 2007, at 8:16 AM, William A. (Andy) Adamson wrote:

>
>
> ---------- Forwarded message ----------
> From: William A. (Andy) Adamson <andros at citi.umich.edu>
> Date: Sep 18, 2007 10:15 AM
> Subject: Re: October Bakeathon: Draft 13 and NFSv4.1 OPEN clientid  
> and other seqid/clientid questions.
> To: lisa.week at sun.com
>
> oops! my mistake! please see below.
>
> On 9/18/07, William A. (Andy) Adamson < andros at citi.umich.edu> wrote:
> Hi
>
> In the interest of interoperabilty in the face of a bit of  
> confusion in draft 13, here is what I propose that the Linux server  
> will do WRT NFSv4.0 sequence IDs and client IDs for NFSv4.1.
>
> CLOSE, LOCK(old lockowner) , LOCKU, OPEN_DOWNGRADE
> - ignore sequence ID

Yes, the Solaris server ignores the seqid in these cases.

>
> LOCK (new lockowner)
> - ignore OPEN client ID and sequence ID

Again, we ignore.

>
> OPEN
> - ignore client ID
> - ignore sequence ID

For us, the behavior on OPEN is currently configurable.  By default,  
we just ignore any seqid and clientid, but are able to change the  
behavior such that it returns NFS4ERR_INVAL if you send us a non-zero  
seqid or clientid.

>
> OPEN is the really confusing one  - I belive the intention in draft  
> 13 is to ignore all client ID and sequence ID for NFSv4.1, and  
> instead derive the client ID from the Session ID, and use the  
> Session sequence ID.
>
>
> and that is what it says....
>
> draft 13 OPEN (section 18.16.4) says
>
> The client ID associated with the owner is
>    not derived from the client field of the owner parameter but is
>    instead the client ID associated with the session on which the
>    request is issued.  If the client ID field of the owner  
> parameter is
>    not zero, the server MUST return an NFS4ERR_INVAL error.
>
> but I don't believe it !!
>
> ^^^^^^^^^^^^^^^^^
> sorry - the above paragraph is fine - my mistake.
>
> I would still like to know what the other implementations are doing...

>
> -->Andy
>
> What are other implementations doing?
>
> Regards
>
> -->Andy
>
> _______________________________________________
> pNFS mailing list
> pNFS at linux-nfs.org
> http://linux-nfs.org/cgi-bin/mailman/listinfo/pnfs

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://linux-nfs.org/pipermail/pnfs/attachments/20070921/25212770/attachment.htm