Requesting Assistance - CentOS4.4 and Ubuntu Install

[AlexanderH]

Good morning,

I have been struggling for the last 24 hours with setup of Zimbra on a VPS environment with 1gb dedicated RAM. In all cases, after the standard installation pains, I was able to get the actual interface up and running just fine. It allows the admin to log in to both admin and the client interface, however no messages can successfully be sent.

I am unaware what the root cause of my problem is at this point, so kindly require further assistance. I am pretty sure there are no conflicts at this point as most all other services have been disabled, as this is a box dedicated for Zimbra, and the standard HTTP has been set to 88. I can access the site just fine at port 88 but sending mail, even to the local admin account that was set up during installation (the only email accont set up on this box) results in an error message after 8 seconds that reads:

-----
A network error has occurred.
msg - csfe service error
code - NETWORK_ERROR
method - ZmCsfeCommand.prototype.invoke
detail - Empty HTTP response
-----

From shell, I see that messages are being sent to the administrator, as it notifies me. They read:

From zimbra@cleo.ewebzone.net Wed May 9 06:40:05 2007
Date: Wed, 9 May 2007 06:40:02 -0500
From: root@cleo.ewebzone.net (Cron Daemon)
To: zimbra@cleo.ewebzone.net
Subject: Cron /opt/zimbra/libexec/zmqueuelog
X-Cron-Env:
X-Cron-Env:
X-Cron-Env:
X-Cron-Env:

postqueue: fatal: Queue report unavailable - mail system is down

---------

From zimbra@cleo.ewebzone.net Wed May 9 06:24:03 2007
Date: Wed, 9 May 2007 06:24:02 -0500
From: root@cleo.ewebzone.net (Cron Daemon)
To: zimbra@cleo.ewebzone.net
Subject: Cron /opt/zimbra/libexec/zmstatuslog
X-Cron-Env:
X-Cron-Env:
X-Cron-Env:
X-Cron-Env:

Could not create the Java virtual machine.

----------


I do believe tomcat is running and is just fine, however. I am able to access the Zimbra interface just fine, can restart tomcat without error, and see intensive java usage in top (which looks normal).

I am most familiar with CentOS and intended on using this distribution, however had no success. I then switched to Ubuntu install. It appears things installed much smoother, but I'm having the same problem.

A records for cleo.ewebzone.net have bee propertly set, as have a mx record on the authorative DNS server for the root domain, ewebzone.net. The name of my zimbra box is obviously cleo.ewebzone.net.

Thank you for your assistance. Please e-mail alex [at] ewebzone.net if you can probide direct assistance. Donations for kind, helpful hands will be given for good karma.

Thanks.

[phoenix]

Well, according to your DNS records externally they point to an EXIM mail server.

Have you disabled any firewall, SElinux and other mail servers on this box? What does 'zmcontrol status' show? Were there any errors during the install? Are there any errors in the logs? Is your hosts file correct, do you have correct DNS A & MX records set? Please post some diagnostics showing that it's working.

[AlexanderH]

Bill,

Thank you for your kind response. I have read your tips given to others, and appreciate you taking the time to help me, as well.

eWebZone.net is a live domain powered by cPanel, with exim mail services. The Zimbra install is server cleo.ewebzone.net. I have created a zone for cleo.ewebzone.net and specified the appropriate a records and mx records. There is no firewall running on cleo at this time.

I have uploaded tmp logs to:
http://www.ewebzone.net/zimbratemp

Of particular interest will probably be:
http://www.ewebzone.net/zimbratemp/zmsetup.log.21848

I appreciate your help, and will do anything that I can to provide you with the most relevant information to make this as quick as possible for you.

Please let me know what I can further provide.

Thank you in advance.
Alex

Originally Posted by phoenix

Well, according to your DNS records externally they point to an EXIM mail server.

Have you disabled any firewall, SElinux and other mail servers on this box? What does 'zmcontrol status' show? Were there any errors during the install? Are there any errors in the logs? Is your hosts file correct, do you have correct DNS A & MX records set? Please post some diagnostics showing that it's working.

[AlexanderH]

Bill,

Thank you for your kind response. I have read your tips given to others, and appreciate you taking the time to help me, as well.

eWebZone.net is a live domain powered by cPanel, with exim mail services. The Zimbra install is server cleo.ewebzone.net. I have created a zone for cleo.ewebzone.net and specified the appropriate a records and mx records. There is no firewall running on cleo at this time.

I have uploaded tmp logs to:
http://www.ewebzone.net/zimbratemp

Of particular interest will probably be:
http://www.ewebzone.net/zimbratemp/zmsetup.log.21848

- My hosts file has been trimmed to the bare minimum while we diagnose this:

127.0.0.1 localhost.localdomain localhost

- A and MX records are in place on server joey.ewebzone.net and pointing to cleo.ewebzone.net

- Installation was a breeze with this distro, and there were no errors or even conflicts given, though I made the http port 88.

- zmcontrol status shows some problems now:
zimbra@cleo:~$ zmcontrol status
Host cleo.ewebzone.net
antispam Running
antivirus Running
ldap Running
logger Running
mailbox Stopped
tomcat is not running
mta Stopped
postfix is not running
snmp Running
spell Stopped
zmapachectl is not running
zimbra@cleo:~$ zmcontrol start
Host cleo.ewebzone.net
Starting logger...FAILED


Starting mailbox...Done.
Starting antispam...Done.
Starting antivirus...Done.
Starting snmp...FAILED


Starting spell...FAILED
Syntax error on line 232 of /opt/zimbra/conf/httpd.conf:
Cannot load /opt/zimbra/httpd-2.0.54/modules/libphp5.so into server: /opt/zimbra/httpd-2.0.54/modules/libphp5.so: undefined symbol: xmlRelaxNGCleanupTypes


Starting mta...Done.
zimbra@cleo:~$ zmcontrol status
Host cleo.ewebzone.net
antispam Running
antivirus Running
ldap Running
logger Running
mailbox Stopped
tomcat is not running
mta Stopped
postfix is not running
snmp Running
spell Stopped
zmapachectl is not running
zimbra@cleo:~$





I appreciate your help, and will do anything that I can to provide you with the most relevant information to make this as quick as possible for you.

Please let me know what I can further provide.

Thank you in advance.
Alex

Originally Posted by phoenix

Well, according to your DNS records externally they point to an EXIM mail server.

Have you disabled any firewall, SElinux and other mail servers on this box? What does 'zmcontrol status' show? Were there any errors during the install? Are there any errors in the logs? Is your hosts file correct, do you have correct DNS A & MX records set? Please post some diagnostics showing that it's working.

[ewebzone]

For some reason, it won't let my reply post be posted to the board:

I will also sign on chat networks if you could help me any way, it'd be appreciated:

AIM: Jabulus
AIM: eWebZone Support
Yahoo: eWebZone
GTalk: AlexHijab

Google email: alexhijab@gmail.com
------------

Bill,

Thank you for your kind response. I have read your tips given to others, and appreciate you taking the time to help me, as well.

eWebZone.net is a live domain powered by cPanel, with exim mail services. The Zimbra install is server cleo.ewebzone.net. I have created a zone for cleo.ewebzone.net and specified the appropriate a records and mx records. There is no firewall running on cleo at this time.

I have uploaded tmp logs to:
http://www.ewebzone.net/zimbratemp

Of particular interest will probably be:
http://www.ewebzone.net/zimbratemp/zmsetup.log.21848

- My hosts file has been trimmed to the bare minimum while we diagnose this:

127.0.0.1 localhost.localdomain localhost

- A and MX records are in place on server joey.ewebzone.net and pointing to cleo.ewebzone.net

- Installation was a breeze with this distro, and there were no errors or even conflicts given, though I made the http port 88.

- zmcontrol status shows some problems now:
zimbra@cleo :~$ zmcontrol status
Host cleo.ewebzone.net
antispam Running
antivirus Running
ldap Running
logger Running
mailbox Stopped
tomcat is not running
mta Stopped
postfix is not running
snmp Running
spell Stopped
zmapachectl is not running
zimbra@cleo:~$ zmcontrol start
Host cleo.ewebzone.net
Starting logger...FAILED


Starting mailbox...Done.
Starting antispam...Done.
Starting antivirus...Done.
Starting snmp...FAILED


Starting spell...FAILED
Syntax error on line 232 of /opt/zimbra/conf/httpd.conf:
Cannot load /opt/zimbra/httpd- 2.0.54/modules/libphp5.so into server: /opt/zimbra/httpd-2.0.54/modules/libphp5.so: undefined symbol: xmlRelaxNGCleanupTypes


Starting mta...Done.
zimbra@cleo:~$ zmcontrol status
Host cleo.ewebzone.net
antispam Running
antivirus Running
ldap Running
logger Running
mailbox Stopped
tomcat is not running
mta Stopped
postfix is not running
snmp Running
spell Stopped
zmapachectl is not running
zimbra@cleo:~$





I appreciate your help, and will do anything that I can to provide you with the most relevant information to make this as quick as possible for you.

Please let me know what I can further provide.

Thank you in advance.
Alex



EDIT - I have since restarted the server and i am able to send outgoing mail. Incoming mail is not being delivered however

[phoenix]

Your hosts file should contain the following:

127.0.0.1 localhost.localdomain localhost
192.168.x.x cleo.ewebzone.net cleo

obviously your LAN IP goes in there.

Make that change and restart zimbra, what shows a running now? Are there emails in your deferred queue? Do they get sent if you flush them?

[AlexanderH]

Forgive the duplicates. I was unaware that approval was required.

Thank you again for your response.

I have previously had that in the hosts file, and was just testing some alternate settings.

Looking at the Zimbra log, I believe I am experiencing the problem as described in:

Mail is delivered to the MTA, but not to the mailbox
http://wiki.zimbra.com/index.php?tit..._Mail_Problems

Analyzing Zimbra logs, I see many postfix/deferred statements.

May 9 06:04:03 cleo postfix/postqueue[9234]: fatal: Queue report unavailable - mail system is down
May 9 06:04:04 cleo zimbramon[7391]: 7391:info: 2007-05-09 06:04:01, STATUS: : Cannot: determine

As mentioned, outgoing messages are being delivered fine, howerver incoming is not, and the message is not failing.

It appears themail server is unable to start:

May 9 06:04:07 cleo postfix/postfix-script: warning: not owned by root: /opt/zimbra/postfix-2.2.9/conf/main.cf
May 9 06:04:07 cleo postfix/postfix-script: starting the Postfix mail system
May 9 06:04:07 cleo postfix/master[20416]: fatal: bind 0.0.0.0 port 25: Address already in use

I also see some connection refused errors, but this is for sendmail oddly:

May 9 09:20:08 cleo sendmail[23886]: l49EK1DQ023886: to=zimbra, ctladdr=zimbra (1000/110),
delay=00:00:04, xdelay=00:00:00, mailer=relay, pri=30286, relay=[127.0.0.1] [127.0.0.1], dsn=4.0.0,
stat=Deferred: Connection refused by [127.0.0.1]

I should point out that I am running a single server for all Zimbra operations.

Can anyone provide feedback from this point forward? Thank you.

[phoenix]

Well, in that case you already have a mailserver running on that system.

It's in your log:

May 9 06:04:07 cleo postfix/master[20416]: fatal: bind 0.0.0.0 port 25: Address already in use

kill it and disable it from starting then start the zimbra mta.

[AlexanderH]

Having determining most of my problems were caused due to resource limitations, I have placed myself on a new VPS with 1.25 dedicated RAM and 4 GB burst.

Now, installation goes through without a hitch. I'm receiving no errors. When I search all logs, I can't find a single error, as opposed to the many I had before.

Prior to my centos install I made sure there were no conflicting services running. I telnetted to the following ports they would not respond: 80, 25, 110, 995 -- all of which were previously giving me problems. I permanently removed apache from the server, as well as sendmail. Again, install went with no problems.

After intsall, doing a zmcontrol status shows all services running but mailbox. The web interface is not up. I can't connect via port 80 or the admin page either. Also tried changing the default http port to 88, with no success.

Telnetting to port 80 or 88 does not connect, but I can see mail services are running properly on 25 and 110.

If I try to start mailbox using zmmailboxctl start it tells me that mysql is already running. I then zmmysqlctl stop and start again, and start mailbox again, and it works fine... this time mailbox says mysql successfully started.

Despite that, doing a zmcontrol status still shows mailbox as offline, as well as tomcat. There are no errors listed in the log for these, however.

Still cant connect to port 80 at all. If i try to zmapachectl start it tells me httpd is already running, and I can see it in top.

Really running into problems here because I'm not getting any error messages.

I was assuming just apache isnt running locally and all mail services are running fine, so I downloaded the desktop account manager and it can't connect to the URL either, and informs me invoke Connection refused.

Kindly advise. Requesting help via IM if possible. Donations for good karms for kind tips.

Please help me end this 36 hour installation.

[AlexanderH]

Again, in summary:

- zmcontrol status shows all online except for mailbox. and it says tomcat is not running. as described in previous post, i have attempted methods of starting individual services and noting the result. no apparent tomcat errors in log.

- Zimbra web server is active (it confirms httpd is already running) but web pages aren't viewable. Telnetting into port shows connection refused. Tested extensively with no success. If I start the standard web server (apache2) after stommping Zimbra services, my html root is displayed perfectly. httpd IS visible in top as an active process. stopping and starting apache, mysql, tomcat has no effect.

- This happens in all of my setups. I have tested various setups prior to installation onto CentOS... with and without services enabled. No matter what I do, I don't get any errors in logs (other than "documents") but have no success in displaying pages or using desktop application successfully.

- it doesnt appear to be a mysql password problem as described often, as i can use mysqladmin just fine.

- not a firewall issue. attempted extensively on Ubuntu and CentOS VPS configurations. Now trying CentOS with no apache success. Previously tried Ubuntu but had extensive Java Virtual Machine issues with errors everywhere in logs. Not the case anymore.


Any feedback is appreciated.