httpd error: Address already in use: AH00072

Question

I have a server, CentOs 6.5 which were installed xampp by a friend of mine. But I thought it's best to install apache, php and mysql from yum. I have it removed simply by using this

#rm -rf /opt/lampp

And I thought it has been completely removed but it didn't happen that way. Maybe I forgot to stop the service httpd. Then I proceed to install Apache 2.4 by following this steps. The error came out when I try to start

service httpd24-httpd start

Starting httpd: (98)Address already in use: AH00072: make_sock: could not bind to address 202.190.185.62:80
no listening sockets available, shutting down
AH00015: Unable to open logs

how do I fix this? Do I have to reinstall the apache by sudo yum install httpd

R Schultz · Answer 1 · 2014-11-26T20:49:41.810

1

To find the process and kill it do the following

netstat -lnp | grep 80 will produce an output something like this.

unix  2      [ ACC ]     STREAM     LISTENING     80   1804/somesoftware     /somepath/and/filename

The 80 is the port number and the 1804 in this case is the process id.

kill -9 1804

Will terminate the process that is taking port 80. The problem is, that when you restart your machine, there is a good chance the software will start over again and you'll have to repeat the process.

Looking at your description of the problem. Try the following first.

service httpd24-httpd stop

You may already be running the apache server from the yum install.

edited Nov 26 '14 at 20:49

answered Nov 26 '14 at 03:13

R Schultz

462
4
12

If you post the output from your netstat and the results of find /etc/rc.d, we may be able to help more. – R Schultz Nov 26 '14 at 03:14
`tcp 0 0 :::80 :::* LISTEN 4032/httpd` – Muhaimin Nov 26 '14 at 03:15
and results of ps -ef | grep 4032 – R Schultz Nov 26 '14 at 03:16
the run `kill -9 4032` and nothing happen. When run again `netstat` it produce another process with another id – Muhaimin Nov 26 '14 at 03:16
`root 11356 10871 0 11:32 pts/0 00:00:00 grep 4032` – Muhaimin Nov 26 '14 at 03:17
There appears to be a parent process restarting it. The results of the ps and the find should help. Do the ps on the process id from netstat before killing it. – R Schultz Nov 26 '14 at 03:17
Let us [continue this discussion in chat](http://chat.stackexchange.com/rooms/18976/discussion-between-r-schultz-and-muhaimin-abdul). – R Schultz Nov 26 '14 at 03:20
`ps -p 4032 -A` ? – Muhaimin Nov 26 '14 at 03:23
You can instead use `fuser 80/tcp` – V13 Oct 02 '15 at 13:25

score 1 · Answer 2 · answered Dec 02 '15 at 16:32

This happened to me when I had a VirtualHost configuration file under custom directory /etc/httpd/sites-enabled that had on top Listen 80

Happens that as it was already listening to this port on httpd.conf, it conflicted, resulting in that error. Removing multiple entries on Listen 80 and having only one at httpd.conf solves it. As I create VirtualHosts listening to ports instead of domains, I always include in the top of the VirtualHost .conf file Listen 12345 the number there corresponding to the adequate port.

score 0 · Answer 3 · edited Nov 26 '14 at 03:14

0

type netstat -lnp | grep 80 and run, you will find which process listen on 80 port, then stop it and start httpd again.

edited Nov 26 '14 at 03:14

PersianGulf

10,728
8
51
78

answered Nov 26 '14 at 02:59

hiworld

11

now the `httpd is stop`. I've tried this before but not sure what to do with the output of netstat – Muhaimin Nov 26 '14 at 03:01
I suspect the running process is from the xampp which I removed it accidentanly – Muhaimin Nov 26 '14 at 03:13

score 0 · Answer 4 · edited Apr 05 '19 at 22:54

0

Except the solution find process running on :80 and kill, then start again,

This error might have if you have multiple "Listen" entries in apache conf file or in any .conf files Included in apache conf file.

edited Apr 05 '19 at 22:54

Rui F Ribeiro

55,929
26
146
227

answered Jun 21 '17 at 13:48

LPince

111
3

score 0 · Answer 5 · answered Jun 21 '17 at 17:07

Quick Workaround

Kill the current instance of the webserver. You may want to change port 80, to 443 if you are using that port instead. After this you should be able to start your instance.

kill -SIGWINCH $(fuser 80/tcp 2>/dev/null | sort -n | awk {'print $1'})

SIGWINCH makes Apache serve any pending requests before shutting down gracefully. For a more forceful shutdown you can change to SIGKILL.

Solution

Identify where you redundant install comes from and remove it, so you can install and rely on your distributions stock package for your webserver.

# fuser 80/tcp 2>/dev/null | grep [0-9] | xargs ps u
USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME 
COMMAND
root      5336  0.0  2.7 421492 27928 ?        Ss   May29   2:47 
/usr/sbin/apache2 -k start
www-data 14390  0.0  0.8 421948  8484 ?        S    16:59   0:00 
/usr/sbin/apache2 -k start
www-data 14395  0.0  0.8 421948  8396 ?        S    16:59   0:00 
/usr/sbin/apache2 -k start
www-data 14396  0.0  0.8 421964  8432 ?        S    16:59   0:00 
...

If you find a path that is in /opt, /usr/local, /opt or similar it is a strong indication that you have a local, manual installation in redundant to the distributions webserver package. Remove it and any start-up scripts that comes with it!

score -1 · Answer 6 · edited Jun 11 '20 at 14:16

-1

Step 1 : netstat -lnp | grep 80

tcp 0 0 :::80 :::* LISTEN 4032/httpd

Step 2 : kill -9 4032

Step 3: Service start httpd

The above simple 3 steps will resolve your issue.

edited Jun 11 '20 at 14:16

Community

1

answered Apr 21 '16 at 09:34

user3588554

1

2

That's just a restatement of the other answers and doesn't add any useful information. – Anthony Geoghegan Apr 21 '16 at 09:45

httpd error: Address already in use: AH00072

6 Answers6

tcp 0 0 :::80 :::* LISTEN 4032/httpd