https://issues.apache.org/bugzilla/show_bug.cgi?id=48169
Rainer Jung changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|
https://issues.apache.org/bugzilla/show_bug.cgi?id=48169
--- Comment #4 from Troy Bowman 2010-01-06 16:51:04 UTC ---
(In reply to comment #3)
> Can you check if the
> http://svn.apache.org/viewvc?rev=892725&view=rev
> fixes the issue?
>
> It uses SOCK_CLOEXEC|FD_CLOEXEC so it should close all so
https://issues.apache.org/bugzilla/show_bug.cgi?id=48169
--- Comment #3 from Mladen Turk 2009-12-21 00:34:01 UTC ---
Can you check if the
http://svn.apache.org/viewvc?rev=892725&view=rev
fixes the issue?
It uses SOCK_CLOEXEC|FD_CLOEXEC so it should close all sockets on fork()
--
Configure bugm
https://issues.apache.org/bugzilla/show_bug.cgi?id=48169
--- Comment #2 from Troy Bowman 2009-11-11 16:24:45 UTC ---
The linux distro is Gentoo. Best distro for people who like to drive
stick-shift. ;)
Thanks to your explanation, I changed the following and the delay indeed
completely disappear
https://issues.apache.org/bugzilla/show_bug.cgi?id=48169
--- Comment #1 from Rainer Jung 2009-11-10 16:10:35
UTC ---
Which Linux do you exactly use?
The cgi process closes the existing backend connections after start (likely
because of CLOEXEC). Since we set SO_LINGER this is done by trying to
