Rick Macklem | 1 Jul 2011 02:10
Picon
Favicon

Re: LOCK and locker

Trond Myklebust wrote:
> On Thu, 2011-06-30 at 10:57 -0400, david.noveck <at> emc.com wrote:
> > That's one of them. The other is that the lokowner information has
> > to stay around even past CLOSE.
> 
> I don't understand this statement.
> 
> * There is already language in RFC3530 that describes how the
> server should deal with a LOCKU replay after CLOSE.
> * We definitely don't want to allow a new LOCK request using an
> old lock stateid after CLOSE.
> 
> So what is the point of keeping the lockowner after CLOSE, and where
> is
> the text that says we need to do so?
> 
I just read Sec. 8.1.7 (of rfc3530) and it seems to say "may" be released.
(And David, my appologies, I see it does define lock_owner, although the definition is
 a bit confusing to me because it uses the term for both open and byte range
 locks.)

However, I agree with Trond in that I don't see a reason to keep it around
after the Close.

Note that, for an open_owner4 string, when the server sees one that it still
has, but the client uses a different open_seqid#, it uses this open_seqid# as the new
initial value and requires an OpenConfirm to verify that it should be the new
one and this isn't an out of order replay.
(When all the open stateids for the open_owner4 string have been Closed.)

(Continue reading)

Rick Macklem | 1 Jul 2011 22:40
Picon
Favicon

Re: LOCK and locker

Thomas Haynes wrote:
> Rick and I have been going over an issue with the following sequence:
> 
> 
> 
> 
> - I'll use labels like O1, L1 for the stateids, to save typing
> Packet#
> 119 open lockfile1279
> 120 ok fh f7ca5653 stateid O1
> 135 open lockfile1279
> 136 ok (should be same fh) stateid O2
> 137 Lock stateid O1 new_lock_owner lock_seqid 0
> 138 ok stateid L1
> 143 LockU fh f7ca5653 stateid L1 lock_seqid 1
> 144 ok
> 147 Close fh f7ca5653 stateid O2
> 148 ok
> 149 Close fh f7ca5653 stateid O1
> 150 ok
> 
> - now on to the next test which creates the file again, etc
>  but does pretty much the same thing
> 159 open lockfile1279
> 160 ok fh f6ca5753 stateid O3
> 175 open lockfile1279
> 176 ok (should be same fh) stateid O4
> 177 Lock f6ca5356 stateid O3 new_lock_owner lock_seqid 0
> 178 ok stateid L2
> 183 LockU fh f6ca5356 stateid L2 lock_seqid 1
(Continue reading)

david.black | 5 Jul 2011 23:02

Re: Quebec and LAYOUT_COMMIT

> > Has anyone volunteered to present something on the LAYOUT_COMMIT issues
> > we are turning into a 5661 errata and a NFSv4.2 update?

Uhm - I guess he means me ;-).

It looks like the change is confined to one paragraph, in section 18.42.3 of RFC 5661:

OLD

   The LAYOUTCOMMIT operation commits changes in the layout represented
   by the current filehandle, client ID (derived from the session ID in
   the preceding SEQUENCE operation), byte-range, and stateid.  Since
   layouts are sub-dividable, a smaller portion of a layout, retrieved
   via LAYOUTGET, can be committed.  The byte-range being committed is
   specified through the byte-range (loca_offset and loca_length).  This
   byte-range MUST overlap with one or more existing layouts previously
   granted via LAYOUTGET (Section 18.43), each with an iomode of
   LAYOUTIOMODE4_RW.  In the case where the iomode of any held layout
   segment is not LAYOUTIOMODE4_RW, the server should return the error
   NFS4ERR_BAD_IOMODE.  For the case where the client does not hold
   matching layout segment(s) for the defined byte-range, the server
   should return the error NFS4ERR_BAD_LAYOUT.

NEW

   The LAYOUTCOMMIT operation commits changes in the layout represented
   by the current filehandle, client ID (derived from the session ID in
   the preceding SEQUENCE operation), and stateid.  As a layout-independent
   operation, LAYOUTCOMMIT commits the entire layout; layout-specific data
   (loca_layoutupdate) may specify a layout subset that requires layout-
(Continue reading)

Spencer Shepler | 5 Jul 2011 23:04
Picon

Re: Quebec and LAYOUT_COMMIT


How does this correlate with what Ricardo captured and passed
by the WG earlier this year?

The text of which can be found here:

http://www.rfc-editor.org/errata_search.php?eid=2751

> -----Original Message-----
> From: nfsv4-bounces <at> ietf.org [mailto:nfsv4-bounces <at> ietf.org] On Behalf Of
> david.black <at> emc.com
> Sent: Tuesday, July 05, 2011 2:02 PM
> To: thomas <at> netapp.com
> Cc: nfsv4 <at> ietf.org
> Subject: Re: [nfsv4] Quebec and LAYOUT_COMMIT
> 
> > > Has anyone volunteered to present something on the LAYOUT_COMMIT
> > > issues we are turning into a 5661 errata and a NFSv4.2 update?
> 
> Uhm - I guess he means me ;-).
> 
> It looks like the change is confined to one paragraph, in section 18.42.3
> of RFC 5661:
> 
> OLD
> 
>    The LAYOUTCOMMIT operation commits changes in the layout represented
>    by the current filehandle, client ID (derived from the session ID in
>    the preceding SEQUENCE operation), byte-range, and stateid.  Since
>    layouts are sub-dividable, a smaller portion of a layout, retrieved
(Continue reading)

david.black | 5 Jul 2011 23:05

Re: Labeled NFS: Per-client subject label functionality - Never Mind

In response to my original posting on this topic, I've seen a lot of discussion, but
no support for adding SET_LABEL to Labeled NFS.  That suggests that this operation is
not needed, at least not for the initial version of Labeled NFS.

OTOH, we do need to make sure that the use of configuration-based information (e.g.,
IP address, VLAN tag) to infer a subject label is explained.

Thanks,
--David

> -----Original Message-----
> From: Black, David
> Sent: Thursday, June 23, 2011 9:30 PM
> To: nfsv4 <at> ietf.org
> Cc: Black, David
> Subject: Labeled NFS: Per-client subject label functionality
> 
> From the call minutes:
> 
> > David provided this taxonomy:
> 
> > Dumb - can't spell Label
> > Label aware - passes/stores object labels, calls back label changes on open files
> > Semi-smart - Implicit or explicit subject label for the client. Could set once.
> > Fully-smart - Subject label per compound - i.e., per users or process on the client
> >
> [... snip ....]
> >
> > We had consensus in the room for David to provide a high overview of
> > how a semi-smart implementation could work, i.e., do we need an op,
(Continue reading)

david.black | 5 Jul 2011 23:13

Re: Quebec and LAYOUT_COMMIT

That's a completely separate and distinct issue.

Thanks,
--David

> -----Original Message-----
> From: Spencer Shepler [mailto:spencer.shepler <at> gmail.com]
> Sent: Tuesday, July 05, 2011 5:05 PM
> To: Black, David; thomas <at> netapp.com
> Cc: nfsv4 <at> ietf.org
> Subject: RE: [nfsv4] Quebec and LAYOUT_COMMIT
> 
> 
> How does this correlate with what Ricardo captured and passed
> by the WG earlier this year?
> 
> The text of which can be found here:
> 
> http://www.rfc-editor.org/errata_search.php?eid=2751
> 
> > -----Original Message-----
> > From: nfsv4-bounces <at> ietf.org [mailto:nfsv4-bounces <at> ietf.org] On Behalf Of
> > david.black <at> emc.com
> > Sent: Tuesday, July 05, 2011 2:02 PM
> > To: thomas <at> netapp.com
> > Cc: nfsv4 <at> ietf.org
> > Subject: Re: [nfsv4] Quebec and LAYOUT_COMMIT
> >
> > > > Has anyone volunteered to present something on the LAYOUT_COMMIT
> > > > issues we are turning into a 5661 errata and a NFSv4.2 update?
(Continue reading)

Rick Macklem | 5 Jul 2011 23:22
Picon
Favicon

SetClientID specifying no callback path

Hi,

I vaguely recall that specifying "0.0.0.0.0.0" for r_addr in a
SetClientID Op was the agreed upon way to specify "no callback path".

Tom thinks it is a NULL (I assume he means zero length) r_addr.

For the 3 servers I have here to test against (Solaris10, Linux/Fedora15
and FreeBSD), either:
   "tcp", "0.0.0.0.0.0"
and
   "", ""   (zero length r_netid and r_addr)

both work to disable callback attempts. (ie. No CB Null is attempted
by the server and no delegations are issued.)

So, what is the preferred way to specify this?

And am I correct to assume that a server will attempt a CB Null to
verify that the callback path is working before issuing a Delegation
to the client?

Thanks in advance for any help with this, rick
_______________________________________________
nfsv4 mailing list
nfsv4 <at> ietf.org
https://www.ietf.org/mailman/listinfo/nfsv4

Spencer Shepler | 5 Jul 2011 23:54
Picon
Favicon

Re: Quebec and LAYOUT_COMMIT


Oh, good that we refer to them as LAYOUTCOMMIT issue. :-)

Having straightened that out.

Why add language specific to 5663 or other non-file layout types
in 5661?  If special handling is needed, presumably it is in the
context of the layout type and can be captured in the RFC for
that layout type.  Therefore, shouldn't the 5661 update be more
generic and then make it clear that additional MUST/SHOULD handling
of LAYOUTCOMMIT will be captured in 5663, etc.?

Spencer

> -----Original Message-----
> From: nfsv4-bounces <at> ietf.org [mailto:nfsv4-bounces <at> ietf.org] On Behalf Of
> david.black <at> emc.com
> Sent: Tuesday, July 05, 2011 2:14 PM
> To: spencer.shepler <at> gmail.com; thomas <at> netapp.com
> Cc: nfsv4 <at> ietf.org
> Subject: Re: [nfsv4] Quebec and LAYOUT_COMMIT
> 
> That's a completely separate and distinct issue.
> 
> Thanks,
> --David
> 
> 
> > -----Original Message-----
> > From: Spencer Shepler [mailto:spencer.shepler <at> gmail.com]
(Continue reading)

Tom Beglin | 5 Jul 2011 23:58
Picon
Favicon

AUTO: Tom Beglin/Tucson/IBM is out of the office. (returning 07/10/2011)

I am out of the office until 07/10/2011.




Note: This is an automated response to your message "nfsv4 Digest, Vol 87, Issue 3" sent on 7/5/11 15:14:31.

This is the only notification you will receive while this person is away.

_______________________________________________
nfsv4 mailing list
nfsv4 <at> ietf.org
https://www.ietf.org/mailman/listinfo/nfsv4
Thomas Haynes | 6 Jul 2011 00:07
Picon

Re: SetClientID specifying no callback path


On Jul 5, 2011, at 4:22 PM, Rick Macklem wrote:

> Hi,
> 
> I vaguely recall that specifying "0.0.0.0.0.0" for r_addr in a
> SetClientID Op was the agreed upon way to specify "no callback path".
> 
> Tom thinks it is a NULL (I assume he means zero length) r_addr.
> 

From a network trace of z/OS NFSC  client to a filer:

Network File System
    [Program Version: 4]
    [V4 Procedure: COMPOUND (1)]
    Tag: setclientid
        length: 11
        contents: setclientid
        fill bytes: opaque data
    minorversion: 0
    Operations (count: 1)
        Opcode: SETCLIENTID (35)
            client
                verifier: 0xc744f405228d7ea8
                id: <DATA>
                    length: 35
                    contents: <DATA>
                    fill bytes: opaque data
            callback
                cb_program: 0x00000000
                cb_location
                    r_netid: <EMPTY>
                        length: 0
                        contents: <EMPTY>
                    r_addr: <EMPTY>
                        length: 0
                        contents: <EMPTY>
            callback_ident: 0x00000000

> For the 3 servers I have here to test against (Solaris10, Linux/Fedora15
> and FreeBSD), either:
>   "tcp", "0.0.0.0.0.0"
> and
>   "", ""   (zero length r_netid and r_addr)
> 
> both work to disable callback attempts. (ie. No CB Null is attempted
> by the server and no delegations are issued.)
> 
> So, what is the preferred way to specify this?
> 
> And am I correct to assume that a server will attempt a CB Null to
> verify that the callback path is working before issuing a Delegation
> to the client?
> 
> Thanks in advance for any help with this, rick
> _______________________________________________
> nfsv4 mailing list
> nfsv4 <at> ietf.org
> https://www.ietf.org/mailman/listinfo/nfsv4

_______________________________________________
nfsv4 mailing list
nfsv4 <at> ietf.org
https://www.ietf.org/mailman/listinfo/nfsv4


Gmane