I've been fiddling with this for some days and can't but assume it's a bug in one of the Gentoo patches to either the kernel or NFS tools: Basically, NFS locking breaks as soon as I enable jumbo frames on both server and client. touch foobar flock foobar ls works fine in my NFS-mounted home with an MTU of 1500. An MTU of 9000 is great for general net throughput so I wanted to use it on both the server and the clients, but the above sequence hangs indefinitely when I try. I'm aware flock() isn't supposed to work correctly with NFS anyway, but all kinds of stuff depends on it at least pretending to. The strange thing is, SuSE 10.1 as a client works fine with jumbo frames, just my Gentoo box doesn't. I tried enabling nfs_debug with sysctl and sniffing the wire with tcpdump and wireshark but with my pretty basic knowledge of NFS workings I didn't spot anything conspicuous other than that lookup(msbethke/foobar) nfs_update_inode(0:18/3424742 ct=1 info=0x6) nfs_fhget(0:18/1081970 ct=1) permission(0:18/1081970), mask=0x4, res=0 seems to be the exchange after which the hang occurs. Our server is running 2.6.18-hardened-r6 and nfs-utils-1.0.12. The clients are mostly SuSE 10.1 boxes with kernel 2.6.16.21-0.21-smp and nfs-utils-1.0.7-36 while my workstation has 2.6.20-gentoo-r6 (was linux-2.6.19-gentoo-r5 before) and the same ns-utils as the server.
-- I prefer encrypted and signed messages. KeyID: FAC37665 Fingerprint: 8C16 3F0m A6FC DF0D 19B0 8DEF 48D9 1700 FAC3 7665
pgpVv5f4MJwd6.pgp
Description: PGP signature