Tomcat/LDAP Failure on 4.5

[russgalleywood]

Hi Guys,

I'm not sure if I will get moaned at for posting htis because i'm not sure if this has already been answered in another thread! If it has I can't find it! I also had a look at the WIKI section as mentioned in the LDAP Issues sticky.

I have never had any problems with LDAP/ Tomcat on any version of Zimbra we have used until now.

Using IMAPsync version 1.182 on ZCS Network Edition 4.5 I used one of the scripts on the WIKI to automate the IMAP transfer. I have used this successfuly on earlier versions of Zimbra.

However when I have come in to work this morning I have found that it failed about half way through saying that the connection to the server it was running on had been refused, this is of course the same physical server that Zimbra is running on.

zmcontrol status told me that Tomcat was not running, I stopped and started zmcontrol and Tomcat started working fine again. I then reran my script and it immediately failed on the first user with the same connection refused message. Restarted the Zimbra machine and tried script again with same result.
Now, when I stop zmcontrol I get this message if I tail the zimbralog;

'Jan 22 11:08:40 servername zimbramon[11260]: 11260:info: Stopping services
Jan 22 11:08:40 servername zimbramon[11260]: 11260:info: Stopping mta
Jan 22 11:08:40 servername postfix/postfix-script: fatal: the Postfix mail system is not running
Jan 22 11:08:40 servername zimbramon[11260]: 11260:info: Stopping spell
Jan 22 11:08:40 servername zimbramon[11260]: 11260:info: Stopping snmp
Jan 22 11:08:40 servername zimbramon[11260]: 11260:info: Stopping antivirus
Jan 22 11:08:40 servername zimbramon[11260]: 11260:info: Stopping antispam
Jan 22 11:08:40 servername zimbramon[11260]: 11260:info: Stopping imapproxy
Jan 22 11:08:42 servername slapd[10027]: conn=22 fd=8 ACCEPT from IP=192.168.0.28:33140 (IP=192.168.0.28:389)
Jan 22 11:08:42 servername slapd[10027]: conn=22 op=0 BIND dn="uid=zimbra,cn=admins,cn=zimbra" method=128
Jan 22 11:08:42 servername slapd[10027]: conn=22 op=0 RESULT tag=97 err=49 text=
Jan 22 11:08:42 servername slapd[10027]: conn=22 fd=8 closed (connection lost)
Jan 22 11:08:42 servername zimbramon[11260]: 11260:info: Stopping mailbox
Jan 22 11:08:42 servername zmtomcatmgr[11435]: threaddump requested
Jan 22 11:08:42 servername zmtomcatmgr[11435]: stale pid 4990 found in /opt/zimbra/log/zmtomcatmgr.pid: No such process
Jan 22 11:08:42 servername zmtomcatmgr[11435]: assuming no other instance is running
Jan 22 11:08:42 servername zmtomcatmgr[11435]: no manager process is running
Jan 22 11:08:44 servername zmtomcatmgr[11443]: stop requested
Jan 22 11:08:44 servername zmtomcatmgr[11443]: stale pid 4990 found in /opt/zimbra/log/zmtomcatmgr.pid: No such process
Jan 22 11:08:44 servername zmtomcatmgr[11443]: assuming no other instance is running
Jan 22 11:08:44 servername zmtomcatmgr[11443]: no manager process is running
Jan 22 11:08:44 servername zimbramon[11260]: 11260:info: Stopping logger
Jan 22 11:08:44 servername zimbramon[11260]: 11260:info: Stopping ldap'

Not sure if this is fine but thought I'd post it just incase.

When I attempt to restart zmcontrol I get this;

'Starting ldap...Done.
FAILED
daemon: bind(7) failed errno=98 (Address already in use)
slap_open_listener: failed on ldap://servername:389
ERROR - failed to start slapd

And in the zimbralog;

Jan 22 11:13:07 servername zimbramon[11697]: 11697:info: Starting services
Jan 22 11:13:07 servername slapd[11733]: @(#) $OpenLDAP: slapd 2.3.21 (Oct 19 2006 13:17:10) $ build@build.liquidsys.com:/home/build/p4/main/ThirdParty/openldap/openldap-2.3.21/servers/slapd
Jan 22 11:13:07 servername slapd[11733]: daemon: bind(7) failed errno=98 (Address already in use)
Jan 22 11:13:07 servername slapd[11733]: slapd stopped.
Jan 22 11:13:07 servername slapd[11733]: connections_destroy: nothing to destroy.
Jan 22 11:13:10 servername slapd[11737]: @(#) $OpenLDAP: slapd 2.3.21 (Oct 19 2006 13:17:10) $ build@build.liquidsys.com:/home/build/p4/main/ThirdParty/openldap/openldap-2.3.21/servers/slapd
Jan 22 11:13:10 servername slapd[11737]: daemon: bind(7) failed errno=98 (Address already in use)
Jan 22 11:13:10 servername slapd[11737]: slapd stopped.
Jan 22 11:13:10 servername slapd[11737]: connections_destroy: nothing to destroy.'

If I now run zmcontrol status is says;

'cannot determine services- - exiting'

I know there are similar message to this on the Forum but not when using IMAPSYNC as far as I can tell.

So, i'm running on CentOS 4.4 (and have not updated it since installing Zimbra), Zimbra version 4.5.0_GA_612.RHEL4_20070115193052, using IMAPSYNC version 1.182.

What am I doing wrong?

Thanks for any help.

Russ

[russgalleywood]

I should also add that this is an upgrade from 4.0.5 version.

[phoenix]

Did you check if LDAP was already running, if it can't start because the addres is in use then it's usually not been shutdown. Try 'zmcontrol stop' and kill any running zimbra services or just a reboot.

[russgalleywood]

Thanks Phoenix,

I tried 'zmcontrol stop' and rebooted, still the same.

Also tried following the Wiki mentioned but am not sure if I am doing this right because I don't get the results on the Wiki!

If I run;

'ps auxww | grep zimbra | grep slapd'

I get;

'zimbra 3173 0.0 0.1 19756 5448 ? Ssl 13:36 0:00 /opt/zimbra/openldap-2.3.21/libexec/slapd -l LOCAL0 -4 -u zimbra -h ldap://servername:389 -f /opt/zimbra/conf/slapd.conf'

If I then run;

'ldap status'

I get nothing back. This means ldap is not running I believe. So then I run;

'killall -TERM slapd'

and next run;

'ps auxww | grep zimbra | grep slapd'

this gives me

'zimbra 4396 0.0 0.0 3956 640 pts/0 S+ 13:58 0:00 grep slapd'

I assume from this that the pid I need to kill is 3956, so I run;

kill -9 3956

But this gives back;

bash: kill: (3956) - No such process

Please take pity on a poor unfortunate fool and explain where I am going wrong as this is most likely the cause of the problem I realise!

Thanks

Russ

[phoenix]

Originally Posted by russgalleywood

'ps auxww | grep zimbra | grep slapd'

this gives me

'zimbra 4396 0.0 0.0 3956 640 pts/0 S+ 13:58 0:00 grep slapd'

When you get to this stage, LDAP is gone. What you're seeing in that output is the 'ps' command itself as you're running it as the zimbra user and searching for ldap.

[russgalleywood]

Ah, so, what next?

[jholder]

Install ZCS.

[russgalleywood]

Sorry?

Do you mean re-install Zimbra?

[russgalleywood]

I have now re-installed 4.5 from scratch on a fresh server and hopefully it won't happen again.

[jholder]

I actually meant to reinstall over your current install. yowsers!