headers
Benny Halevy | 3 Jan 2011 15:03
Favicon

Re: Questions on pNFS

[trimmed mailing lists on the Cc]

On 2010-12-30 04:48, Zhangyu wrote:
> Hi,
> 
> I have successfully installed the pNFS and it works well.

Great!

> However, I am interested in the parallel access control mechanism of pNFS.
> I tried to edit the same file on two different clients,
> I found there wasn't any byte-range lock as specified in the ietf-nfsv4-draft protocol.

Correct.

> How is the parallel access control implemented? Which source code file does it located in? 
> 

There is none by definition.
The pNFS feature is orthogonal to synchronization primitives so
different clients that need to serialize their access should use locking
to do that.

Some servers may implement synchronization at the pnfs layout level,
but the linux server has no such feature.

> Additionally, I found that when writing a file to the DS, network flow is higher both
> on MDS and DS. Does it mean that the data is delivered through MDS?
> I think the data should be written directly to DS, is it right?
(Continue reading)

Permalink | Reply | 2 comments |
headers
Benny Halevy | 3 Jan 2011 15:21
Favicon

Re: layoutcommits and file layout

On 2010-12-17 01:07, Christoph Hellwig wrote:
> On Thu, Dec 16, 2010 at 11:21:21AM -0500, Matt W. Benjamin wrote:
>> Hi,
>>
>> We have a files implementation which wants to receive LAYOUTCOMMIT when a client is finished with a
layout.  It was my clear understanding from rfc5661 that we could expect this behavior.
> 
> Care to post it to the list?
> 

I don't know what Matt's server is doing but the fundamental problem is
manifested with extending a file with parallel DS writes.
Assuming that the DS writes are executed in arbitrary order,
exposing the file length before LAYOUTCOMMIT can cause
a concurrent reader to read a hole.  Although locking can
solve this case, day-to-day applications that work well over
local filesystem and legacy NFS may break because of this.

Benny
_______________________________________________
nfsv4 mailing list
nfsv4 <at> ietf.org
https://www.ietf.org/mailman/listinfo/nfsv4
Permalink | Reply | 4 comments |
headers
david.noveck | 6 Jan 2011 22:20

Re: Questions on pNFS

I'm very confused by this interchange.

> > I found there wasn't any byte-range lock as specified in the ietf-nfsv4-draft protocol.

> Correct.

Huh?

The nfsv4 protocol does have byte-range locks.  It is optional whether they are advisory or mandatory, but
they do exist.

Could Benny mean that they don't work in his server?  It doesn't seem so because he advises Zhangyu to use
locking. 

The only form of mandatory locking in NFSv4 are share reservations.  Waiting to edit your file may be drag but
it is better than turning it into a mess.

I've never seen an editor that allowed multiple users to edit a file and used byte-range locks to
synchronize.  It seems like a fun project as long as there was no delivery schedule.

-----Original Message-----
From: nfsv4-bounces <at> ietf.org [mailto:nfsv4-bounces <at> ietf.org] On Behalf Of Benny Halevy
Sent: Monday, January 03, 2011 9:03 AM
To: Zhangyu
Cc: linux-nfs; nfsv4
Subject: Re: [nfsv4] Questions on pNFS

[trimmed mailing lists on the Cc]

On 2010-12-30 04:48, Zhangyu wrote:
(Continue reading)

Permalink | Reply | 1 comment |
headers
Spencer Shepler | 7 Jan 2011 04:28
Picon
Favicon

NFSv4 Interim meeting (Feb 18-19) - Action: submit agenda topics


Given the date/location proximity of Usenix FAST and Connectathon '11,
we will be holding the NFSv4 interim working group meeting on
Feb 18th/19th.

Netapp has offered to host the interim meeting at their site in Sunnyvale.
Specific details are forthcoming.

If possible, a one day interim meeting would most agreeable.
However, it seems as though we have enough agenda material for
two days.  To validate duration, please submit agenda items asap.

The intent of this interim meeting is to take the time and have
detailed discussions about current protocol problems, existing 
proposals with issue closure and detailed actions for next steps.
The first priority will be to close on issues regarding NFSv4.0
and NFSv4.1.  Secondary will be NFSv4.2 proposals. 

As a reminder, this interim meeting; it is not a replacement for
the next full IETF meeting in Prague.  We will be meeting there
as well.

I will submit for an official IETF interim meeting announcement
at the end of next week; this will allow for and agenda to
be constructed and to gather final details of meeting location.

Thanks,
Spencer

_______________________________________________
(Continue reading)

Permalink | Reply | 5 comments |
headers
david.noveck | 7 Jan 2011 20:42

Re: NFSv4 Interim meeting (Feb 18-19) - Action: submit agendatopics

The major item I have is the discussion of a range of pNFS questions
with regard to the layout synchronization and sequencing.  In many case,
the discussion on the working group list ended but it wasn't always
clear if we had a clear spec-directed resolution (everything is OK,
errata, or a some other clarification mechanism).  In other cases, it
seemed the discussion was suspended due to people getting tired.

I've promised to come up with detailed list of issues.  Let me renew
that promise and commit to deliver a draft to the working group by 2/1,
as way of motivating me (and trying to create some time to do it on my
schedule).

I think we should not get involved in interminable discussions (easy to
say).  I would allocate an hour for the discussion.  A second session of
15-20 minutes would be good (on the second day or the afternoon if we
are a one-day meeting).  That second session would be devoted to making
sure we have the agreement we think we have and recruiting volunteers to
write up the resolutions agreed to.

I'm sure there will be a range of RFC3530bis issues.  I'll let Tom speak
to those.

-----Original Message-----
From: nfsv4-bounces <at> ietf.org [mailto:nfsv4-bounces <at> ietf.org] On Behalf
Of Spencer Shepler
Sent: Thursday, January 06, 2011 10:28 PM
To: nfsv4 <at> ietf.org
Subject: [nfsv4] NFSv4 Interim meeting (Feb 18-19) - Action: submit
agendatopics
(Continue reading)

Permalink | Reply | 1 comment |
headers
Spencer Shepler | 7 Jan 2011 20:51
Picon
Favicon

Re: NFSv4 Interim meeting (Feb 18-19) - Action: submit agendatopics


This sounds great and thanks for the commitment for 2/1.  That will
allow for pre-work by those in attendance...

Spencer

> -----Original Message-----
> From: david.noveck <at> emc.com [mailto:david.noveck <at> emc.com]
> Sent: Friday, January 07, 2011 11:42 AM
> To: Spencer Shepler; nfsv4 <at> ietf.org
> Subject: RE: [nfsv4] NFSv4 Interim meeting (Feb 18-19) - Action: submit
> agendatopics
> 
> The major item I have is the discussion of a range of pNFS questions with
> regard to the layout synchronization and sequencing.  In many case, the
> discussion on the working group list ended but it wasn't always clear if
> we had a clear spec-directed resolution (everything is OK, errata, or a
> some other clarification mechanism).  In other cases, it seemed the
> discussion was suspended due to people getting tired.
> 
> I've promised to come up with detailed list of issues.  Let me renew that
> promise and commit to deliver a draft to the working group by 2/1, as way
> of motivating me (and trying to create some time to do it on my schedule).
> 
> I think we should not get involved in interminable discussions (easy to
> say).  I would allocate an hour for the discussion.  A second session of
> 15-20 minutes would be good (on the second day or the afternoon if we are
> a one-day meeting).  That second session would be devoted to making sure
> we have the agreement we think we have and recruiting volunteers to write
> up the resolutions agreed to.
(Continue reading)

Permalink | Reply | 0 comments |
headers
sfaibish | 7 Jan 2011 21:40

Re: NFSv4 Interim meeting (Feb 18-19) - Action: submit agenda topics

Spencer,

These were the topics that we discussed internally at the BAT in Boston
that I propose to discuss openly at the interrim meeting:
1. pNFS file servers implementation strategy
2. pNFS over IPv6
3. Performance related issues of NFSv4.1/pNFS
4. LAYOUTCOMMIT for pNFS file
5. Courtesy Locks in NFSv4.0
6. Secure NFS in 4.x
7. Multi-layout pNFS server support
8. FedFS imlementation issues
9. Discussion on our general testing strategy and new test suite  
(replace/extend cthon)

This is not a prioritized list but a list as it was schedulled at BAT. It  
may be
that some of the items are not of interest for all but the developers so  
it is
up to you to decide if they are all relevant to the WG. Thank you for your
consideration.

/Sorin

On Thu, 06 Jan 2011 22:28:01 -0500, Spencer Shepler  
<sshepler <at> microsoft.com> wrote:

>
> Given the date/location proximity of Usenix FAST and Connectathon '11,
> we will be holding the NFSv4 interim working group meeting on
(Continue reading)

Permalink | Reply | 2 comments |
headers
Spencer Shepler | 10 Jan 2011 08:29
Picon
Favicon

Re: NFSv4 Interim meeting (Feb 18-19) - Action: submit agenda topics


Thanks for the suggestions, Sorin.

Is there anyone interested in championing these
issues either through agenda material or an I-D (if appropriate)?

Spencer

> -----Original Message-----
> From: sfaibish [mailto:sfaibish <at> emc.com]
> Sent: Friday, January 07, 2011 12:41 PM
> To: Spencer Shepler; nfsv4 <at> ietf.org
> Subject: Re: [nfsv4] NFSv4 Interim meeting (Feb 18-19) - Action: submit
> agenda topics
> 
> Spencer,
> 
> These were the topics that we discussed internally at the BAT in Boston
> that I propose to discuss openly at the interrim meeting:
> 1. pNFS file servers implementation strategy 2. pNFS over IPv6 3.
> Performance related issues of NFSv4.1/pNFS 4. LAYOUTCOMMIT for pNFS file
> 5. Courtesy Locks in NFSv4.0 6. Secure NFS in 4.x 7. Multi-layout pNFS
> server support 8. FedFS imlementation issues 9. Discussion on our general
> testing strategy and new test suite (replace/extend cthon)
> 
> This is not a prioritized list but a list as it was schedulled at BAT. It
> may be that some of the items are not of interest for all but the
> developers so it is up to you to decide if they are all relevant to the
> WG. Thank you for your consideration.
>
(Continue reading)

Permalink | Reply | 1 comment |
headers
Thomas Haynes | 10 Jan 2011 19:01
Picon

NFSv4 Interim meeting (Feb 18-19) - Action: Reserve a spot

I'm trying to see if I need to get a room for upto 25 or a larger venue.

If this were a regular working group meeting, I could make an accurate
guess, but since it is juxtaposed with other events and in the Bay Area ....

Please respond directly to me if you plan on attending.

Note that we won't turn anyone away, but by responding early, you get
to influence how many latecomers get to stand. :->
_______________________________________________
nfsv4 mailing list
nfsv4 <at> ietf.org
https://www.ietf.org/mailman/listinfo/nfsv4
Permalink | Reply | 0 comments |
headers
Thomas Haynes | 10 Jan 2011 19:29
Picon

Re: NFSv4 Interim meeting (Feb 18-19) - Action: submit agenda topics


On Jan 9, 2011, at 11:29 PM, Spencer Shepler wrote:


Thanks for the suggestions, Sorin.

Is there anyone interested in championing these
issues either through agenda material or an I-D (if appropriate)?

Spencer




5. Courtesy Locks in NFSv4.0

This will fall under the 3530bis discussion, which I am willing to own.

 3.
Performance related issues of NFSv4.1/pNFS 

 9. Discussion on our general
testing strategy and new test suite (replace/extend cthon)

From the BAT discussions, I felt that both of these were not issues at
the IETF level. I'd prefer we push these off until the Cthon talks.

I.e., they were more tactical discussions on interoperability issues.

1. pNFS file servers implementation strategy 

Wasn't this one more of a survey?


 7. Multi-layout pNFS
server support 8. 

And this one as well?

While I think they both have an impact on the documents, they might also
be better suited to talks at Cthon?


_______________________________________________
nfsv4 mailing list
nfsv4 <at> ietf.org
https://www.ietf.org/mailman/listinfo/nfsv4
Permalink | Reply | 0 comments |

Gmane