Re: nfs write speed performance... still

[Antoine Jacoutot <ajacoutot_(_at_)_lphp_(_dot_)_org>]

On Thursday 25 November 2004 10:31, you wrote:
> Adding the following rule in /etc/pf.conf should sort this out.
> =================================
> # scrub and clear Don't Fragment bit (NFS)
> scrub in all no-df
> =================================

This was already configured that way.
But the issue is between OpenBSD clients and server. Linux writes to OpenBSD 
NFS server are more than fine.

> Of course you may not be using pf and the filesystem may not be on a Linux
> box but in the absence of more information we're all in the dark.

???
Haven't you read all the thread. I provided a _lot_ of information.
Anyway, it seems that the problem is known but no fixes are known or planned 
for now since there're other priorities.

Regards,

Antoine