[SOLVED] Another LDAP failure while upgrading...

[Klug]

Currently trying to upgrade an OSE Ubuntu 32bit from 6.0.4 to 6.0.7.

ZCS was running fine (I was able to stop/start it without issue and made a backup).

I launched the install process.
It starts correctly (remove old packages, install new ones, upgrade the LDAP content), then stop LDAP server.

When it's supposed to restart the LDAP server (to get the installed packages), it fails.

Here's the zmsetup.log :

Sat Jul 17 16:45:44 2010 Getting installed packages
Sat Jul 17 16:45:44 2010 Getting local config ldap_url
Sat Jul 17 16:45:45 2010 zimbra_server_hostname contained in ldap_url checking ldap status
Sat Jul 17 16:45:45 2010 Checking ldap status....
Sat Jul 17 16:45:45 2010 *** Running as zimbra user: /opt/zimbra/bin/ldap status
Sat Jul 17 16:45:45 2010 not running.
Sat Jul 17 16:45:45 2010 Running bdb db_recover...
Sat Jul 17 16:45:45 2010 *** Running as zimbra user: /opt/zimbra/bdb/bin/db_recover -h /opt/zimbra/data/ldap/hdb/db
db_recover: file unknown has LSN 4/6219077, past end of log at 4/4343802
db_recover: Commonly caused by moving a database from one database environment
db_recover: to another without clearing the database LSNs, or by removing all of
db_recover: the log files from a database environment
Sat Jul 17 16:45:45 2010 done.
Sat Jul 17 16:45:45 2010 Running zmldapapplyldif...
Sat Jul 17 16:45:45 2010 *** Running as zimbra user: /opt/zimbra/libexec/zmldapapplyldif
IO::Socket::INET: connect: Connection refused at /opt/zimbra/libexec/zmldapapplyldif line 122.
Sat Jul 17 16:46:25 2010 failed.
Sat Jul 17 16:46:25 2010 Checking ldap status....
Sat Jul 17 16:46:25 2010 *** Running as zimbra user: /opt/zimbra/bin/ldap status
Sat Jul 17 16:46:26 2010 not running.
Sat Jul 17 16:46:26 2010 Starting ldap...
Sat Jul 17 16:46:26 2010 *** Running as zimbra user: /opt/zimbra/bin/ldap start
Failed to start slapd.  Attempting debug start to determine error.
hdb_db_open: database "": db_open(/opt/zimbra/data/ldap/hdb/db/id2entry.bdb) failed: Invalid argument (22).
backend_startup_one (type=hdb, suffix=""): bi_db_open failed! (22)
bdb_db_close: database "": alock_close failed

Sat Jul 17 16:47:02 2010 failed with exit code: 256.
Sat Jul 17 16:48:39 2010 checking isEnabled zimbra-core
Sat Jul 17 16:48:39 2010 zimbra-core not in enabled cache
Sat Jul 17 16:48:39 2010 enabled packages
Sat Jul 17 16:48:39 2010 zimbra_server_hostname contained in ldap_url checking ldap status
Sat Jul 17 16:48:39 2010 Checking ldap status....
Sat Jul 17 16:48:39 2010 *** Running as zimbra user: /opt/zimbra/bin/ldap status
Sat Jul 17 16:48:39 2010 not running.
Sat Jul 17 16:48:39 2010 Running bdb db_recover...
Sat Jul 17 16:48:39 2010 *** Running as zimbra user: /opt/zimbra/bdb/bin/db_recover -h /opt/zimbra/data/ldap/hdb/db
Sat Jul 17 16:48:39 2010 done.
Sat Jul 17 16:48:39 2010 Running zmldapapplyldif...
Sat Jul 17 16:48:39 2010 *** Running as zimbra user: /opt/zimbra/libexec/zmldapapplyldif
IO::Socket::INET: connect: Connection refused at /opt/zimbra/libexec/zmldapapplyldif line 122.
Sat Jul 17 16:49:19 2010 failed.
Sat Jul 17 16:49:19 2010 Checking ldap status....
Sat Jul 17 16:49:19 2010 *** Running as zimbra user: /opt/zimbra/bin/ldap status
Sat Jul 17 16:49:20 2010 not running.
Sat Jul 17 16:49:20 2010 Starting ldap...
Sat Jul 17 16:49:20 2010 *** Running as zimbra user: /opt/zimbra/bin/ldap start
Failed to start slapd.  Attempting debug start to determine error.
hdb_db_open: database "": db_open(/opt/zimbra/data/ldap/hdb/db/id2entry.bdb) failed: Invalid argument (22).
backend_startup_one (type=hdb, suffix=""): bi_db_open failed! (22)
bdb_db_close: database "": alock_close failed

Sat Jul 17 16:49:57 2010 failed with exit code: 256.

db_recover is not able to help; I tried it by hand too.

I also tried to restore /opt/zimbra/data/ldap from the backup (the one running OK), the same happens: upgrade starts but after first ldap stop, it's unable to start again.

Any idea?

[Klug]

I seem to have fixed it using the basic method...

Dump the backup to an ldif file (slapcat) and then load the data (slapadd).

[Klug]

Just for the record, it happened once again (NE upgrade 6.0.6 to 6.0.10).

db_recover did not help.
I did a slapadd of last night backup to solve the problem.

/opt/zimbra/openldap/sbin/slapadd -F /opt/zimbra/data/ldap/config -qcv -l /opt/zimbra/backup/sessions/full-20110108.000004.030/ldap/ldap.bak