On my CentOS 5.2 x86 (kernel: 2.6.18) box I'm running Litespeed 4.0.9 and using IPv6. When I gracefully restart, Litespeed is unable to rebind the IPv6 addresses until the next graceful restart after all old processes shut down.
I'm unsure how to handle this. Is this a problem with my kernel? With the webserver? Or my configuration? My bet is the kernel or the webserver, but hopefully you can help me clear that up. 
Code:
2009-08-01 18:59:51.520 ERROR HttpListener::start(): Can't listen at address IPv6 DLP: Address already in use!
2009-08-01 18:59:51.520 ERROR HttpServer::addListener(IPv6 DLP) failed to create new listener
2009-08-01 18:59:51.520 ERROR [config:server:listener:IPv6 DLP] failed to start listener on address [2607:f0d0:1002:xx]:80!
2009-08-01 18:59:51.826 ERROR HttpListener::start(): Can't listen at address IPv6 Patron: Address already in use!
2009-08-01 18:59:51.826 ERROR HttpServer::addListener(IPv6 Patron) failed to create new listener
2009-08-01 18:59:51.826 ERROR [config:server:listener:IPv6 Patron] failed to start listener on address [2607:f0d0:1002:xx]:80!