NetBSD-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Very poor NFS I/O





On 1/03/2026 1:33 am, BERTRAND Joël wrote:
MJ a écrit :

[...snip...]

     Only nfsd and kernel :

legendre# ps auwx | grep ' D'
root       23258 68.9  0.1  614028   8312 ?      Dsl  12:14PM    0:36.32
/usr/sbin/nfsd -n 128
root           0  0.1  1.1       0 182868 ?      DKl  12Feb26  807:12.02
[system]


Yikes. Have you tried running the server with far less threads? I,
personally, can't see the benefit of such a high amount of threads.
This is probably inviting contention.

My usual rule-of-thumb was (2 * cores ) + 2. Not scientific, just
historically, "it worked".

So, have you tried the server running -n 16 or -n 20 or similar.

	I have tried with -n16. nfs server seems to run better, but client
receives a lot of "nfs server not responding". Linux can handle with
this message (even in diskless configuration), but FreeBSD can crashes.


Off-topic, but interestingly, OpenBSD sets a limit for nfs servers (threads) of 20, default of 4:
#define	MAXNFSDCNT	20
#define	DEFNFSDCNT	 4



Perhaps you also need to provide the client side configurations. While there's issues with the server-side, perhaps some of that can be mitigated by tweaking the clients.

Timeout settings due to soft mounts? Read/write buffers too small?
I generally do hard client mounts with rsize/wsize set to 32K.

Regards
Max.


Home | Main Index | Thread Index | Old Index