I'm running "ab2 -n10000 -c1" to stress my lighttp + scgi + python
configuration.
but I receive time outs (???)

If I'm doing a "netstat -a", I can see such tings (+4000 lines):
tcp        0      0 localhost:4000          localhost:59313         TIME_WAIT
tcp        0      0 localhost:4000          localhost:51089         TIME_WAIT
tcp        0      0 localhost:4000          localhost:58545         TIME_WAIT
tcp        0      0 localhost:4000          localhost:50321         TIME_WAIT
tcp        0      0 localhost:4000          localhost:58801         TIME_WAIT
tcp        0      0 localhost:4000          localhost:50577         TIME_WAIT
tcp        0      0 localhost:4000          localhost:58033         TIME_WAIT

What's the problem ? lighttpd ? scgi ?
Where to look first ?

My config is:
- lighttpd-1.4.10
- scgi-1.10
- server.event-handler = "linux-sysepoll" (I'm on gentoo with kernel-2.6.15)

lighttpd was compiled with (Gentoo USE):
+bzip2, +doc, +fam, +fastcgi, +gdbm, +lua, +memcache +pcre, +rrdtol, +ssl,
+webdav, +xattr
-ipv6, -ldap, -minimal, -mysql, -php, -test

Thanks

PS: same post on the lighttpd mailing list (for follow-up)
http://thread.gmane.org/gmane.comp.web.lighttpd/3611



--
William: http://www.opensource4you.com