migrate from old scalix box to new one

[kabalah]

Hi there, i'm going to migrate a scalix 10 community edition from a suse box to a red had 4 es box. during the migration, the mailnode from the old box needs to be changed on the new box. is there a procedure i can use to migrate all the users, pdls and mail? i haven't seen anything relating to this on the forum or the knowlege base...will the ldapsync work to do this?
thanks for any help
jim

[kabalah]

well, the old box has crashed for good....so it looks like i won't need to migrate anything.
we are just recreateing the environment from scratch on the new server.

what would be the procedure for this if i have to do it in the future...?
jim

[Valerion]

There's 2 ways of doing it:

1) For each user on the old box, do a omcpoutu of his mailstore to a file
2) Install Scalix on the new machine
3) Create all the users on the new server in the correct mailnode
4) Transfer the files created in step 1
5) Do a omcpinu for every user

This will retain the mailbox contents, but will lose the mailstore configuration and the MAPI data in each mailbox (especially who it is shared out to).

The second way is to

1) Create the scalix user on the new mail server
2) Copy the entire /var/opt/scalix from the old server to the new one. It is critical you preserve the file and directory ownerships and permissions.
3) Install Scalix on the new machine, preserving the mailstore when asked
4) Create the new mailnode with omaddmn
5) Use ommodu and ommodpdl to change the mailnode each user and PDL (group) is in
6) Delete the mailnode with omdelmn
7) If there is permissions errors in the mailstore you can use omcheck to correct them.

Is there any specific reason you want to change the mailnode?

[kabalah]

after taking the advanced admin class, i just wanted to shorten the mail node to get rid of the hostname, domain....and to shorten it to just the domain. it didn't really matter in the long run though, the raid controller died and wrote crap over the disks so, we haven't been able to recover any of the old data. the backups were also causing this box to panic....so, the down side is that the users lost two months of data, but the upside is that we now have a properly built with redhat 4es/lvm2 server with 4 times a day snapshots copyied to a hot standby server which gets backed up nightly. i just wish they would have let me do it sooner.
thanks for your help
jim