We have a redhat 7.3 nfs client that doesn't work well with an irix
6.5.14m nfs server.  When a gnome user logs in, things are fine for
about 5 minutes, but consistently, the network gets ground to dust after
about 5 minutes or so.

We tried reverting to nfs v2, and that didn't help.

The guy in charge of these machines is going to downgrade the redhat box
to 7.2 to see if that fixes it.  These two machines were fine with each
other prior to the 7.3 upgrade, when the redhat was 7.2.

Is anyone else seeing this sort of thing?  Does anyone have any
suggestions?

Oh, BTW, ethereal reports -tons- of nearly identical fragmented UDP
packets, not that it's easy to use ethereal when the subnet's getting
beat up so badly.

I suggested trying a 7.2 kernel on the 7.3 box, but the guy in charge of
the box feels the enduser is too impatient to wait for that, and will
grow angry if it doesn't work.

-- 
Dan Stromberg                                               UCI/NACS/DCS

Attachment: msg80995/pgp00000.pgp
Description: PGP signature

Reply via email to