On Fri, May 06, 2005 at 02:38:18PM +0200, Jose Luis Marin wrote: > > Hi there, > > We have a moderately-loaded NFS server running Debian/Sarge with the > stock kernel 2.6.8-10-amd64-k8 (the current Debian binary deb). It > serves around 10 NFS client connections from a Solaris 8 box. For > what it's worth, the exported filesystems are also being shared > simultaneously via Samba to windows clients (although I doubt this is > the problem). > > We're getting kernel GPFs once every 3 weeks or so, and it seems to be > on process nfsd. This is the case at least on the two occasions where > there was a trace left in the logs. See below. > > I checked my ram with memtest86+, and I even changed the eth card to a > more stable e1000, but the problem must be somewhere else... Any ideas?
Are you able to test this against the 2.6.11 kernels that are in unstable? -- Horms -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
