Performance Difference based on .... ?

Hello list,

I have come to a dead-end...
How come on a freshly mounted file-system, compiling takes considerably less time than on subsequent runs ? I mean, this snip of shell:
---snip---
for i in 1 2 3
do
time ./apachectl configtest
done
---snip---
2:17.23 on the first run, while it takes
3:21.92 on the second and
3:20.78 on the third run.

The network is 1gb ethernet, with jumbo frames 9000 and tests were run both on 4096 and 8192 rsize/wsize with more or less similar results. Client is 2x AthlonMP 1800+, server is 2x AthlonMP 2200+, kernel on both is 2.4.26+Trond's linux-2.4.26.NFS.ALL patch,
mount is udp - essentially so, because this is meant to be a high availability setup.
The only real limitation is the rate at which the server can feed data to the client, since cpu is 31% on the first and 20% on the second and third runs.

Any ideas ? Any other info I can give that might help ?

Regards,
Thanos Chatziathanassiou

-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs