Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

FC7->FC8 upgrade yields apache error: (98)Address already in use: make_sock: could not bind to address [::]:80

0 views
Skip to first unread message

Eric H.

unread,
Nov 14, 2007, 12:03:00 PM11/14/07
to
This may be an Apache issue and if so I'll take my questions elsewhere
but I upgraded from FC7 to FC8 and now for some odd reason when I
restart apache I get this error:

"(98)Address already in use: make_sock: could not bind to address [::]:80
(98)Address already in use: make_sock: could not bind to address 0.0.0.0:80
no listening sockets available, shutting down
Unable to open logs"


But I can kill the running httpd and apache will start without a
problem...no warnings whatsoever. I never got this error prior to
upgrading.

I've changed absolutely nothing about my httpd.conf but during testing I
did note that the default httpd.conf installed doesn't cause the restart
error. I'm guessing that's because the default httpd.conf doesn't setup
virtualhosts, etc. but it might be significant. Apache's test shows no
errors in the .conf file. I don't have any files in the conf.d directory
so there are no conflicts there.


Any ideas?

Eric H.

unread,
Nov 14, 2007, 12:50:14 PM11/14/07
to
I just noted that I've been using "service httpd restart" to restart
Apache and when I tried "apachectl restart" it works fine so ostensibly
this is a Fedora issue.

Eric H.

unread,
Nov 14, 2007, 1:49:42 PM11/14/07
to
/etc/init.d/httpd relies on a static reference to the httpd.pid and I
had moved mine into another directory in httpd.conf. Seems to be time
they derive that from the httpd.conf since it's not guaranteed to be in
a static location and this is a VERY non-obvious bug to hunt out.

Ivan Marsh

unread,
Nov 14, 2007, 2:02:36 PM11/14/07
to
On Wed, 14 Nov 2007 13:49:42 -0500, Eric H. wrote:

> /etc/init.d/httpd relies on a static reference to the httpd.pid and I
> had moved mine into another directory in httpd.conf. Seems to be time
> they derive that from the httpd.conf since it's not guaranteed to be in
> a static location and this is a VERY non-obvious bug to hunt out.

Moving data from where it's expected can hardly be called a bug.

Doc it hurts when I do this.

--
I told you this was going to happen.

Eric H.

unread,
Nov 14, 2007, 2:16:37 PM11/14/07
to
Ivan Marsh wrote:
> Moving data from where it's expected can hardly be called a bug.
>
> Doc it hurts when I do this.


If the path was hard coded in the apache binary then I might agree with
you but it's not; it's in httpd.conf for a reason so my assertion
stands, it's a bug with the "service" command.

Ivan Marsh

unread,
Nov 14, 2007, 3:00:58 PM11/14/07
to
On Wed, 14 Nov 2007 14:16:37 -0500, Eric H. wrote:

> Ivan Marsh wrote:
>
>> Moving data from where it's expected can hardly be called a bug.
>

> If the path was hard coded in the apache binary then I might agree with
> you but it's not; it's in httpd.conf for a reason so my assertion
> stands, it's a bug with the "service" command.

I will admit that if the init script is getting all it's other
configuration data from the conf file it might as well get the pid as
well. I still wouldn't call it a bug... poorly written and redundant?
Sure.

0 new messages