Nfs nlockmgr timed out

thanks for reading me and helping me !

I got a NFS problem. It's not a permanent problem it comes and goes.
Sometimes (very often), the clients using the mounted partitions are frozen when they're trying to open a new program like eclipse or kde.
I think it's a lock problem due to maybe a network problem or optimization problem, but I don't know what to do, I tried a lot of thing but I failed….

The ubuntu version

cat /etc/issue
Ubuntu 10.04.4 LTS \n \l

When Everything is OK :

Every 2.0s: rpcinfo -u 10.10.10.2 nlockmgr              Wed Jan  9 13:50:41 2013
rpcinfo: RPC: Program/version mismatch; low version = 1, high version = 4program
 100021 version 1 ready and waiting
program 100021 version 2 is not available
program 100021 version 3 ready and waiting
program 100021 version 4 ready and waiting

When the problem happens

Every 2.0s: rpcinfo -u 10.10.10.X nlockmgr Wed Jan 9
13:58:39 2013
rpcinfo: RPC: Timed out
program 100021 version 0 is not available

No problem with others like mountd or status

Server log :

Jan 9 14:02:25 dallas kernel: [95801.650031] statd: server rpc.statd
not responding, timed out Jan 9 14:02:25 dallas kernel:
[95801.650044] lockd: cannot monitor topalbum

client log :

kernel: lockd: server 10.10.10.X not responding, timed out

Jan 9 13:22:24 topalbum kernel: [6464607.366966] lockd: server
fileserver OK Jan 9 13:23:24 topalbum kernel: [6464667.367109] lockd:
server fileserver OK Jan 9 13:24:24 topalbum kernel: [6464727.520031]
lockd: server fileserver not responding, still trying Jan 9 13:24:24
topalbum kernel: [6464727.520115] lockd: server fileserver not
responding, still trying

avahi-daemon[3890]: Invalid query packet.

I tried to increase the RPCNFSDCOUNT in /etc/default/nfs-kernel-server using

echo "512" > /proc/fs/nfsd/threads

I tried to mount client partitions with no lock, tcp….

I tested the network performance with iperf, got 936 Go/s

I got statd enabled

server is up to date

I am exhausted…