httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
Subject Re: Apache 2.0 on
Date Fri, 16 Feb 2001 16:51:37 GMT

> Thanks, Ryan, but I had *lots* of help.  
> 1.3 came up at 00:00:01 PST and grabbed port 80.  I didn't do it (3 AM
> here).  Brian didn't do it, but about an hour and a half later he
> noticed a residual 2.0 parent running, not listening to a TCP port, so
> he killed it.  I don't think you did it either, Ryan.  Smells like a
> cron job to me, because of the time.  
> I don't understand how port 80 got taken without killing 2.0.  If 2.0
> was killed, it didn't come down cleanly.  Maybe there's clues in the
> logs.
> Other than that, I don't see any reason why 2.0 couldn't have stayed in
> production.

I have a couple of thoughts.  First of all, let's get 2.0 back up on  We can find the cron job that killed us later.  I believe
this was the rotatelogs script.

I believe we were able to re-grab the port because Apache 2.0 grabbed the
IPv6 port 80, and 1.3 grabbed the IPv4 80.  If they were both started at
the some time, is it possible for there to be a race condition between
grabbing the IPv6 port and IPv4 port?

Just a question.  Would anybody mind if I stopped 1.3 and brought up 2.0
in the next hour?  I'll watch it all day to see what is happening.

Ryan Bloom               
406 29th St.
San Francisco, CA 94131

View raw message