NFS server flooded and vfs.nfsd.tcphighwater

Hello.

 From time to time, on an NFS4 server of mine, I see this message:
> nfsd server cache flooded, try increasing vfs.nfsd.tcphighwater

It's typically when I installworld on a client (src and obj being on 
this server) and it's repeated over and over.

Nothing nasty seems to happen (perhaps only a slowdown???) and the lack 
of documentation leaves me wondering what to do (if anything) with this.

man page says:
> If the server has stopped servicing clients and has generated a console
>      message like "nfsd server cache flooded...", the value for
>      vfs.nfsd.tcphighwater needs to be increased.  This should allow the
>      server to again handle requests without a reboot.  Also, you may want to
>      consider decreasing the value for vfs.nfsd.tcpcachetimeo to several
>      minutes (in seconds) instead of 12 hours when this occurs.

Well, for a start it doesn't really look like the server stopped 
servicing...
Then there's no indication of what value should be reasonable. In fact 
it's not even clear what unit we are talking about (bytes? kB? some 
unspecified count?).

If vfs.nfsd.tcphighwater had a starting value I could think about 
doubling it, raising it tenfold or whatever; alas it's default value is 0.

Some forums show people used some apparently random value like 300000, 
while admitting they had no idea what they were doing.

Any pointer?

  bye & Thanks
	av.