Port 2087/2083 daily issues

[JeremyL]

I will try and catch the log at the same time it stops working next time. One thing I have noticed is when I restart CPanel I get

Stopping Cpanel3 services: [ OK ]
Stopping Cpanel3 log services: [ OK ]
Stopping Cpanel3 Chat services: [ OK ]
Stopping Melange Chat services: [ OK ]
Stopping InterChange services: [FAILED]
Stopping Cpanel3 ssl services: [ OK ]
Stopping Web Host Manger services: [ OK ]
Stopping Cluster Management services: [ OK ]
Starting Cpanel3 services: [ OK ]
Starting WebMail services: [ OK ]
Starting pop3 services: [ OK ]
Starting Cpanel3 Chat services: [ OK ]
Starting Melange Chat services: [ OK ]
Starting Cpanel3 ssl services: [ OK ]
Starting Web Host Manager services: [ OK ]
Starting Cluster Management services: [ OK ]

The Interchange is not supposed to stop or start but it does start. I have done what it said in these forums to stop interchange since we don\'t use it but CPanel still seems to think it is still trying to work. Could this be an issue?

[JeremyL]

OK here is the output from /var/log/messages when the cron does it\'s deal

Dec 13 22:15:01 dixie cpanel3: cpaneld shutdown succeeded
Dec 13 22:15:01 dixie cpanel3: cpanellogd shutdown succeeded
Dec 13 22:15:01 dixie cpanel3: entropychat shutdown succeeded
Dec 13 22:15:01 dixie cpanel3: melange shutdown succeeded
Dec 13 22:15:01 dixie cpanel3: interchange shutdown failed
Dec 13 22:15:01 dixie stunnel[16472]: Received signal 15; terminating
Dec 13 22:15:01 dixie stunnel[16474]: Received signal 15; terminating
Dec 13 22:15:01 dixie cpanel3: stunnel shutdown succeeded
Dec 13 22:15:01 dixie cpanel3: whostmgrd shutdown succeeded
Dec 13 22:15:02 dixie cpanel3: clustermgrd shutdown succeeded
Dec 13 22:15:02 dixie cpaneld: ==> Loading Config ....product myname adminuser basename version root engineroot engine enginepl cycle docroot access_log stats_log port RS VALIASDIR VFILTERDIR
Dec 13 22:15:02 dixie cpaneld: ==> Loading Config ....product myname adminuser basename version root engineroot engine enginepl cycle docroot access_log stats_log port RS VALIASDIR VFILTERDIR
Dec 13 22:15:02 dixie cpaneld: ==> cpaneld(logger) 4.0 Starting ( processor system).....
Dec 13 22:15:02 dixie cpaneld:
Dec 13 22:15:02 dixie cpaneld: ==> cpaneld 4.0 Starting ( processor system).....
Dec 13 22:15:02 dixie cpanel3: cpaneld startup succeeded
Dec 13 22:15:02 dixie webmaild: ==> Webmaild 3.5.0 Server Starting.....
Dec 13 22:15:02 dixie cpanel3: webmaild startup succeeded
Dec 13 22:15:02 dixie cppop: Address in use at bin/cppop.pl line 57.
Dec 13 22:15:02 dixie cpanel3: cppop startup succeeded
Dec 13 22:15:02 dixie entropychat: Content-type: text/html
Dec 13 22:15:02 dixie entropychat:
Dec 13 22:15:02 dixie entropychat: entropychat 0.2.5 started
Dec 13 22:15:02 dixie entropychat: <br>Listening on port 2084...
Dec 13 22:15:02 dixie cpanel3: entropychat startup succeeded
Dec 13 22:15:03 dixie cpanel3: startmelange startup succeeded
Dec 13 22:15:03 dixie startstunnel: ==> Starting SSL tunnel...
Dec 13 22:15:03 dixie startstunnel: stunnel: no process killed
Dec 13 22:15:03 dixie stunnel[17574]: Using \'127.0.0.1.2082\' as tcpwrapper service name
Dec 13 22:15:03 dixie stunnel[17574]: stunnel 3.18 on i686-pc-linux-gnu PTHREAD+LIBWRAP
Dec 13 22:15:03 dixie stunnel[17576]: Using \'127.0.0.1.2086\' as tcpwrapper service name
Dec 13 22:15:03 dixie stunnel[17576]: stunnel 3.18 on i686-pc-linux-gnu PTHREAD+LIBWRAP
Dec 13 22:15:03 dixie cpanel3: startstunnel startup succeeded
Dec 13 22:15:03 dixie whostmgrd: ==> Web Host Manager 3.9.0 Server Starting.....
Dec 13 22:15:03 dixie cpanel3: whostmgrd startup succeeded
Dec 13 22:15:03 dixie clustermgrd: ==> Cluster Manager 3.5.0 Server Starting.....
Dec 13 22:15:04 dixie cpanel3: clustermgrd startup succeeded
Dec 13 22:16:11 dixie proftpd[17363]: dixie.svsecure.net (dixie.svsecure.net[127.0.0.1]) - FTP login timed out, disconnected.
Dec 13 22:17:31 dixie kernel: eth0: card reports no resources.
Dec 13 22:18:25 dixie kernel: eth0: card reports no resources.

[feanor]

Nothing too strange...... do you have anything besides portsentry installed that could be yanking ports / firewalling to some extent? Something strange I saw when skimming through was that cppop.pl error, address already in use (means port already has a service *or something* bound to it)

Anything else you\'ve installed on this box, that would be in the genre of portsentry/tripwire, any insane ipchaining / iptabling going on in the backround?

Also, the cpanel restart services script seldom works well with interchange. Build this into the end of your cron:

/usr/local/cpanel/bin/startinterchange &


Word.

[JeremyL]

No I haven\'t added any software that cpanel didn\'t install itself and I haven\'t messed with IPChains

[moronhead]

[quote:f077ae23f5] What is in /var/log/messages in the same ballpark of time as when port 2087 allegedly stops listening? [/quote:f077ae23f5]REFUSED by libwrap. That\'s what it says.

[xidica]

Just FYI, I had an issue with portsentry preventing access to the secure ports while the non-secure were running and stunnel was also running :

root@host [/etc/portsentry]# netstat -l --program | grep stunnel
tcp 0 0 *:imaps *:* LISTEN 29792/stunnel-4.04l
tcp 0 0 *:2083 *:* LISTEN 29792/stunnel-4.04l
tcp 0 0 *op3s *:* LISTEN 29792/stunnel-4.04l
tcp 0 0 *:2087 *:* LISTEN 29792/stunnel-4.04l
tcp 0 0 *:2096 *:* LISTEN 29792/stunnel-4.04l

Simple cpanel restart didn't fix it.

After :
/etc/rc.d/init.d/portsentry stop
/etc/rc.d/init.d/cpanel restart
/etc/rc.d/init.d/portsentry start

Everything is working.

The permanent fix for this is to add the ports to the exclude list so that portsentry will ignore these ports and never bind to them.

in /etc/portsentry/portsentry.conf change :
ADVANCED_EXCLUDE_TCP="113,139" <- default setting from my experience
to :
ADVANCED_EXCLUDE_TCP="113,139,2083,2087,2096"

restart portsentry :
/etc/rc.d/init.d/portsentry restart

Hope this helps someone, and I have no idea why these ports aren't excluded by default. My guess is that portsentry doesn't mess with the non-secure as they don't contain as much gibberish(at least as far as the tcp wrapper is concerned)

[phpcoder1]

I'm having trouble with this, and this is still outputting this whenever I try to restart portsentry/cpanel...

Starting cPanel SSL services: [ FAILED ]

And nothing anybody said would work...

[xidica]

I'm having trouble with this, and this is still outputting this whenever I try to restart portsentry/cpanel...

Starting cPanel SSL services: [ FAILED ]

And nothing anybody said would work...

Are there any errors in the log? Have you checked to see what is in the file /usr/local/cpanel/etc/stunnel/default/stunnel.conf ?
Is it looking for cpanel.pem or mycpanel.pem ? If so does the file exist? i've seen a few issues where it's looking for one, but the other is correct. I've usually just made a symlink so if it looks for either it will find it(as I'm not sure if cpanel upcp's rewrite this conf file). Any unusual error messages in the logs?

[phpcoder1]

What should I do? Logs say nothing about SSL. It's looking for /usr/local/cpanel/etc/cpanel.pem. cPanel.pem has RSA/key stuff... tell me step-by-step what to do by PMing me.