You could make the backup database a slave to your primary database, so that it is constantly being updated. then just switch your config when your primary fails.

-Rex

On Jun 4, 2009, at 11:00 AM, mukim pathan wrote:

Hi,

Thank you for your inputs, Buchan. Actually what I am trying to do is to recreate the database if my present system fails due to any reason. Since ldap takes a lot of time in recovery for my system(it took 9 hours for 40 million entries and 1200 log files), I am trying to stage up another machine in the mean time as quickly as possible.I tried backing up and then reloading, but it seems to be taking a lot of time in generating the backup. So please let me know any faster solution with which I can recreate my database.

Regards, Mukim Pathan

On Thu, Jun 4, 2009 at 7:49 PM, Buchan Milne <bgmilne@staff.telkomsa.net

...

wrote:

On Thursday 04 June 2009 08:58:21 mukim pathan wrote:

...

Hi, I am trying to create new ldap database on new machine and I want to transfer full ldap database from old machine to new machine. Can I

just

...

transfer all the database files (.bdb,_db,alock) and expect new

database to

...

be at the same stage of old database??

This won't work if: -The version of OpenLDAP is different -The database library version is older on the target -The architectures differ (e.g. x86 vs x86_64).

Even if it does work (none of the above conditions apply), it's still not a good idea to copy the database files directly (as it is non-trivial to get a consistent copy).

As covered in the documentation, the supported method for backing up and restoring data is via slapcat to export, and slapadd to import.

Regards, Buchan

mukim pathan wrote:

Hi Rex,

How do I do that? If there are any documentation about it then please direct me to that. Thank you in advance.

Regards, Mukim Pathan

On Thu, Jun 4, 2009 at 8:36 PM, Rex Roof <rex@wccnet.edu mailto:rex@wccnet.edu> wrote:

You could make the backup database a slave to your primary database,
so that it is constantly being updated.
then just switch your config when your primary fails.

-Rex

Hello,

I believe the answer is here > http://www.openldap.org/doc/admin24/replication.html To be more specific, you will have to take only some parts of Syncrepl proxy. This could give you better look at things > http://www.zytrax.com/books/ldap/ch7/

And also, our configs (for ilustration):

--- master ---- # Indices to maintain index objectClass eq

# Indices for SyncRepl index entryCSN,entryUUID eq

## These indexes are included to support calls such as ## getpwuid( ), getpwnam( ), and ## getgrgid( ). index cn,uid eq index uidNumber eq index gidNumber eq

# Replication - go and get me overlay syncprov syncprov-checkpoint 100 10 syncprov-sessionlog 100

# Let the replica DN have limitless searches limits dn.exact="cn=syncuser,dc=domain,dc=tld" time.soft=unlimited time.hard=unlimited size.soft=unlimited size.hard=unlimited --- master ----

--- slave (replica) ---- overlay syncprov

syncrepl rid=002 provider=ldaps://master.domain.tld:636/ binddn="cn=syncuser,dc=domain,dc=tld" bindmethod=simple credentials=clearTextPswd searchbase="dc=domain,dc=tld" # persistent connection type=refreshAndPersist retry="5 5 300 +"

# reference updates to master updateref ldaps://master.domain.tld:636 --- slave (replica) ----

Have a nice evening, Zdenek