Our web server was just moved to a Debian slink box today.
We get a lot of our content onto the box using NFS, with the web server
acting as the NFS server, the content generation host (Solaris 2.6)
mounting the web tree.
Unfortunately, the nfs-server stuff (2.2beta37-1) keeps
failing every few minutes. Specifically, rpc.nfsd dies, with errors
logged like:
Nov 18 12:40:19 silk nfsd[482]: strange write req from vil.rap.ucar.edu: count
4294967295 len 0
Nov 18 12:41:52 silk nfsd[523]: strange write req from spider.rap.ucar.edu:
count 8192 len 5284
Nov 18 12:42:57 silk nfsd[523]: strange write req from sat1.rap.ucar.edu: count
4294967295 len 4294967295
Nov 18 12:43:25 silk nfsd[523]: non-standard errno: 14 (Bad address)
Nov 18 12:43:48 silk last message repeated 3 times
Kernel version is 2.1.128, with kernel NFS disabled in favor
of the userland stuff. Enabling kernel-based NFS changes nothing.
Using 2.0.36 changes nothing. The Debian version is slink, up-to-date
as of this morning.
Help! Thanks... ;)
--
Tres Hofmeister <[EMAIL PROTECTED]> http://www.rap.ucar.edu/staff/tres/
Research Applications Program National Center for Atmospheric Research
