> This week (probably wednesday) I'll put both my netfilter hook statistic
> patch, and enabled kernel profiling, onto a production box (the transproxy
> thing from the bucket occupation analysis). Right now I have totally
> undersized bucket count on that machine (7168 buckets for 10 times
> the tuples), so I'll first measure the "accidental long list walk"
> situation, and then retry with a suitable bucket size.
Before somebody get the wrong idea: the machine I mentioned, serves
as a squid proxy for over 3000 narrowband dialup users (all web
traffic), and it has no performance problems at all with that.
For all I know, any optimization we may make regarding netfilter,
won't make the squids on that box work perceivably better.
I have permanent average and median service time monitoring to prove
or disprove this assertion :-)
best regards
Patrick
|
