Cannot start Zimbra M2

[tbullock]

This is the error I get when I try and start Zimbra after installing the lates Open Source release:

[root@webmail bin]# service zimbra start
Host webmail.avmax.ca
Starting ldap...Done.
ERROR: service.FAILURE (system failure: getDirectContext) (cause: javax.naming.CommunicationException webmail.avmax.ca:389)


This is the error I get when I try and start ldap manually as the zimbra user:

[zimbra@webmail bin]$ ldap start
db_recover: Finding last valid log LSN: file: 1 offset 28
/opt/zimbra/openldap/libexec/slapd: error while loading shared libraries: libbind.so.3: cannot open shared object file: No such file or directory

Any ideas?

[KevinH]

Your system is out of date. So run yum update (or yum install bind-libs) to bring that up to date.

[tbullock]

[zimbra@webmail bin]$ ./zmcontrol start
Host webmail.avmax.ca
Starting ldap...Done.
ERROR: service.FAILURE (system failure: getDirectContext) (cause: javax.naming.CommunicationException webmail.avmax.ca:389)

OK. I have done that. But I am still getting the message above. I have modifed my /etc/hosts and done the following:

delete /opt/zimbra/openldap-data/* (leave the directory)
Make sure all zimbra processes are stopped
su - root
touch /opt/zimbra/.newinstall (Note the '.' in the filename)
/opt/zimbra/libexec/zmsetup.pl

Still no luck.

I am running Fedora Core 3.

Any more ideas?

[KevinH]

Try to start slapd in debug mode. Look at this post for more info:

Upgrade M1 -> M2? .... Yes!!

[woyakj]

Hello. I tried to follow all of the above steps and the ones in the other post but to no avail (I am still getting the exact same error when I do a 'zmcontrol start'). When i directly run '/opt/zimbra/bin/ldap start' after doing the recovery it gives me:

db_recover: Finding last valid log LSN: file: 1 offset 43437
db_recover: Recovery starting from [1][43296]
db_recover: Recovery complete at Fri Nov 18 11:57:47 2005
db_recover: Maximum transaction ID 80000002 Recovery checkpoint [1][43437]
Started slapd: pid 13220

and it starts ldap, but when i do a '/opt/zimbra/bin/zmcontrol start' it gives me the same error as mentioned earlier.

Any ideas? We are a mid-sized public school district and are really looking forward to being able to test the new functionality and decide if Zimbra is the right solution for us. It's even wierder because the M1 release worked without any problems. Thanks for any help you can provide.

[KevinH]

Did you remove iptables? zmiptables -u If you had M1 on the machine then the iptables rules may still be there which we don't use any more.

[Billy]

[root@ns3 ~]# su zimbra
[zimbra@ns3 root]$ su root
Password:
[root@ns3 ~]# zmiptables -u
[root@ns3 ~]# exit
exit
[zimbra@ns3 root]$ exit
exit
[root@ns3 ~]# su zimbra
[zimbra@ns3 root]$ zmcontrol start
Host ns3.xxx.com
ERROR: service.FAILURE (system failure: getDirectContext) (cause: javax.naming.C ommunicationException ns3.xxx.com:389)
[zimbra@ns3 root]$ ldap start
db_recover: Finding last valid log LSN: file: 1 offset 6080366
db_recover: Recovery starting from [1][6080225]
db_recover: Recovery complete at Fri Nov 18 22:05:53 2005
db_recover: Maximum transaction ID 800002ec Recovery checkpoint [1][6080366]
Started slapd: pid 2546
[zimbra@ns3 root]$ zmcontrol start
Host ns3.interalab-dns.com
Starting antispam...Done.
Starting antivirus...Done.
Starting mailbox...Done.
Starting mta...Done.
Starting spell...Done.
[zimbra@ns3 root]$

[agnOstos]

Congrats!

[marcmac]

zmlocalconfig -e ldap_is_master=true will force ldap to start, though the upgrade script should have done that.

[Billy]

After it appeared to start, I checked the status and found that antivirus and mta had not actually started. Eventually I found a conflict between the latest libdb-4.3.so and the db.h in sleepycat that was looking for something around version 4.2.52.

I installed compat-db-4.2.52-2.i386.rpm and everything started running.

(thanks to previous post) I also had to force ldap to start on startup with:

zmlocalconfig -e ldap_is_master=true

I'm still checking to see if it is all working.