tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: NFS vs jumbograms?

(2013/06/04 5:50), David Holland wrote:
On Mon, Jun 03, 2013 at 11:37:29AM -0400, Mouse wrote:
  > At $JOB, we have two i386 machines with wm interfaces connected
  > back-to-back with a short patch cable and a /30 subnet of 192.168.  One
  > is NFS-serving some disk space to the other over this link.  They are
  > running 4.0.1 (with a few tweaks of mine, but nothing touching
  > sys/nfs); while it's fallen off the end of official support, I thought
  > someone might happen to recall enough to be useful....

nfs hasn't changed much, so if the problem is in nfs (rather than in
netinet) it's probably still in -current.

not that this actually helps.

Also note that there are a fair number of cases in nfs where lost
packets result in timeouts that don't get handled properly causing
things to wedge. If you haven't yet, try using tcp mounts as that
tends to avoid exercising some of those cases.

 I sometimes heard that wm had some problems with jumbo MTU, but I've
never checked whether it really has a bug or not. I suspect the bug
is not in NFS stack but in wm.

 I don't use jumbo MTU, so usually don't test the function when I add
the support of a new chip. Sorry :(
If someone show me a reproduceable configuraton, I'll try to fix it.


                SAITOH Masanobu (

Home | Main Index | Thread Index | Old Index