public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* nfs NULL procedure packets
@ 2008-12-03 18:11 Russell Miller
  2008-12-04 18:45 ` J. Bruce Fields
  0 siblings, 1 reply; 2+ messages in thread
From: Russell Miller @ 2008-12-03 18:11 UTC (permalink / raw)
  To: linux-kernel

Hi again, I was hoping someone here could answer another question.  I
apologize if these are easy questions...

So a couple of days ago I was troubleshooting yet another NFS problem
(not the non-one I described a few days ago).  I sent a 1G file over
to a server, did a TCP capture, and analyzed the results.

I noticed a lot of NULL procedure packets.  A *lot* of them.  In some
cases, one every 10 microseconds, and there were about 400,000 of them
I captured.

This seems like a lot of network traffic for very little reason.

Is this normal behavior, or something I should investigate further?
It seems to me like pinging the NFS server once every 10 microseconds
is a major waste of bandwidth.

--Russell

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: nfs NULL procedure packets
  2008-12-03 18:11 nfs NULL procedure packets Russell Miller
@ 2008-12-04 18:45 ` J. Bruce Fields
  0 siblings, 0 replies; 2+ messages in thread
From: J. Bruce Fields @ 2008-12-04 18:45 UTC (permalink / raw)
  To: Russell Miller; +Cc: linux-kernel

On Wed, Dec 03, 2008 at 10:11:26AM -0800, Russell Miller wrote:
> Hi again, I was hoping someone here could answer another question.  I
> apologize if these are easy questions...
> 
> So a couple of days ago I was troubleshooting yet another NFS problem
> (not the non-one I described a few days ago).  I sent a 1G file over
> to a server, did a TCP capture, and analyzed the results.
> 
> I noticed a lot of NULL procedure packets.  A *lot* of them.  In some
> cases, one every 10 microseconds, and there were about 400,000 of them
> I captured.
> 
> This seems like a lot of network traffic for very little reason.
> 
> Is this normal behavior, or something I should investigate further?
> It seems to me like pinging the NFS server once every 10 microseconds
> is a major waste of bandwidth.

It's either a bug, or you're misreading the traces--how are you
identifying these as NULL procedures?  Also, are they NULL NFS calls, or
some other rpc protocol?  And are you using krb5?

--b.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2008-12-04 18:45 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-12-03 18:11 nfs NULL procedure packets Russell Miller
2008-12-04 18:45 ` J. Bruce Fields

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox