Problem starting LDAP

[PhishKiller]

I a similar problem, but it's not a bing error 99.

I've read over the LDAP wiki page and really, it's been that much help.

This is what I am seeing:

[zimbra@localhost ~]$ sudo /opt/zimbra/openldap/libexec/slapd -d7 -l LOCAL0 -4 
-u zimbra -h dap://mail.our-private-domain.com:389 -f /opt/zimbra/conf/slapd.conf
@(#) $OpenLDAP: slapd 2.3.37 (Aug 30 2007 22:58:59) $
        build@build-rhel5-64:/home/build/p4/main/ThirdParty/openldap/openldap-2.3.37.7/servers/slapd
daemon_init: dap://mail.our-private-domain.com:389
daemon_init: listen on dap://mail.our-private-domain.com:389
daemon_init: 1 listeners to open...
ldap_url_parse_ext(dap://mail.our-private-domain.com:389)
daemon: listen URL "dap://mail.our-private-domain.com:389" parse error=3
slapd stopped.
connections_destroy: nothing to destroy.
[zimbra@localhost ~]$

As far as I can tell, my /etc/hosts file is just fine (follows)

[root@localhost ~]# more /etc/hosts
# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1       localhost.localdomain localhost
---.13.100.12   mail.our-private-domain.com mail
::1             localhost6.localdomain6 localhost6
[root@localhost ~]#

DNS is resolving the domain without any problems.

[root@localhost ~]# host mail.our-private-domain.com 
mail.our-private-domain.com has address ---.13.100.12

I checked all of the servers logfiles for anything to do with ldap or zimbra and this is all I found:

secure:Nov  8 17:48:37 localhost sudo:   zimbra : TTY=pts/5 ; PWD=/opt/zimbra ; USER=root ; COMMAND=/opt/zimbra/openldap/libexec/slapd -d7 -l LOCAL0 -4 -u zimbra -h dap://mail.our-private-domain.com:389 -f /opt/zimbra
secure:Nov  8 17:55:53 localhost sudo:   zimbra : TTY=pts/10 ; PWD=/opt/zimbra ; USER=root ; COMMAND=/opt/zimbra/openldap/libexec/slapd -d7 -l LOCAL0 -4 -u zimbra -h dap://mail.our-private-domain.com:389 -f /opt/zimbra/conf/slapd.conf

secure:Nov  8 17:54:30 localhost su: pam_unix(su-l:session): session opened for user zimbra by root(uid=0)
secure:Nov  8 17:55:46 localhost su: pam_unix(su-l:session): session opened for user zimbra by root(uid=0)
secure:Nov  8 17:55:53 localhost sudo:   zimbra : TTY=pts/10 ; PWD=/opt/zimbra ; USER=root ; COMMAND=/opt/zimbra/openldap/libexec/slapd -d7 -l LOCAL0 -4 -u zimbra -h dap://mail.our-private-domain.com:389 -f /opt/zimbra/conf/slapd.conf

I'm out of ideas and the ones I've tried from the wiki docs, other than disabling sendmail, only seem to lead me in circles.

What am I missing here and/or what do I need to try next?

[phoenix]

You should not have attached this to a thread that's marked as [Solved], it will get missed.

I'd suggest you disable ipv6 on your zimbra server. Do you have any firewall or SElinux? If so, you should disable them. Is this an install problem or an already installed system? Some details of operating system and Zimbra version might help. Was there anything in the install logs that indicate an error?

Let's start with a check of your DNS, do the following:

host `hostname`  <-- type as-is with backticks not single quotes
dig domain.com mx
dig domain.com any

and post the results.

[PhishKiller]

running host `hostname` indicated the the hostname was still set to localhost. I've fixed that and have restarted the system.

The DNS and MX records are OK but I'll post additional info after fixing the hostname and attempting the restart of LDAP. IF it's still not working I'll post the dns mx and other info.

THX!

[PhishKiller]

Fixed the hostname but I'm still receiving the same error trying to start ldap

Ran hostname, now it returns the IP and hostname as expected

[root@mail ~]# host `hostname`
mail.our-private-domain.com has address ---.13.100.12

Here is the dig on our root domain's MX record:

[root@mail ~]# dig  our-private-domain.com MX

; <<>> DiG 9.3.3rc2 <<>> our-private-domain.com MX
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 39465
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 4, AUTHORITY: 3, ADDITIONAL: 7

;; QUESTION SECTION:
;our-private-domain.com.          IN      MX

;; ANSWER SECTION:
our-private-domain.com.   3600    IN      MX      10 mail.another-domain.com.
our-private-domain.com.   3600    IN      MX      30 vw1.another-domain.com.
our-private-domain.com.   3600    IN      MX      40 rs4.another-domain.com.
our-private-domain.com.   3600    IN      MX      60 rs2.another-domain.com.

;; AUTHORITY SECTION:
our-private-domain.com.   3600    IN      NS      dns1.another-domain.com.
our-private-domain.com.   3600    IN      NS      post.another-domain.com.
our-private-domain.com.   3600    IN      NS      host1.another-domain.com.

;; ADDITIONAL SECTION:
mail.another-domain.com.   600     IN      A       ---.14.127.210
vw1.another-domain.com.    600     IN      A       ---.23.208.117
rs4.another-domain.com.    600     IN      A       ---.144.186.16
rs2.another-domain.com.    600     IN      A       ---.144.192.114
dns1.another-domain.com.   600     IN      A       ---.14.127.213
post.another-domain.com.   600     IN      A       ---.10.150.125
host1.another-domain.com.  600     IN      A       ---.91.87.141

;; Query time: 1 msec
;; SERVER: ---.14.202.213#53(---.14.202.213)
;; WHEN: Fri Nov  9 15:04:56 2007
;; MSG SIZE  rcvd: 301

Ran the dig on the MX record we'll be using the mx record on the domain itself (our-private-domain.com) is currently running mission critical mail serves we plan to replace with the system being provisioned with the Zimbra server, it's MX record dig follows:

[root@mail ~]# dig testmail.our-private-domain.com mx

; <<>> DiG 9.3.3rc2 <<>> testmail.our-private-domain.com mx
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 8862
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 3, ADDITIONAL: 4

;; QUESTION SECTION:
;testmail.our-private-domain.com. IN      MX

;; ANSWER SECTION:
testmail.our-private-domain.com. 3600 IN  MX      10 mail.our-private-domain.com.

;; AUTHORITY SECTION:
our-private-domain.com.   3600    IN      NS      host1.another-domain.com.
our-private-domain.com.   3600    IN      NS      dns1.another-domain.com.
our-private-domain.com.   3600    IN      NS      post.another-domain.com.

;; ADDITIONAL SECTION:
mail.our-private-domain.com. 3600 IN      A       ---.13.100.12
dns1.another-domain.com.   600     IN      A       ---.14.202.213
post.another-domain.com.   600     IN      A       ---.101.150.125
host1.another-domain.com.  600     IN      A       ---.191.87.141

;; Query time: 1 msec
;; SERVER: ---.14.202.213#53(---.14.127.213)
;; WHEN: Fri Nov  9 14:54:34 2007
;; MSG SIZE  rcvd: 202

Checking the hostname.domain we'll be using

[root@mail ~]# dig mail.our-private-domain.com any

; <<>> DiG 9.3.3rc2 <<>> mail.our-private-domain.com any
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 2002
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0

;; QUESTION SECTION:
;mail.our-private-domain.com.      IN      ANY

;; AUTHORITY SECTION:
com.                    900     IN      SOA     a.gtld-servers.net. nstld.verisign-grs.com. 1194648842 1800 900 604800 900

;; Query time: 13 msec
;; SERVER: ---.14.202.213#53(---.14.202.213)
;; WHEN: Fri Nov  9 14:54:49 2007
;; MSG SIZE  rcvd: 115

I'm not sure how the MX records come into play with LDAP. LDAP has never been on our technology roadmap, so this is new territory.

I attempted to run zmsetup again, and this time I recieved the 99 error. Information follows:

The system will be modified - continue? [No] Yes
Operations logged to /tmp/zmsetup.log.3403
Initializing ldap...Failed to start slapd.  Attempting debug start to determine error.
daemon: bind(7) failed errno=99 (Cannot assign requested address)
slap_open_listener: failed on ldap://mail.our-private-domain.com:389

I saw this in the zmsetup.log file:

checking isEnabled zimbra-ldap
zimbra-ldap is enabled
Initializing ldap...

Is it possible to simply disable zimbra-ldap? We're trying to evaluate this as a replacement for an aging and convoluted sendmail system and it's been 2 days of sytem provisioning and Zimbra install already.

[PhishKiller]

In the end, a completed re-install seemed to fix the problem.

My impression was (it being a couple weeks since the first re-install and just running a re-install again this morning) that it was all due to the DNS and our MX record not being a perfect match for the hostname we were using.

Hopefully this is helpful for others.