Disaster Recovery SAC access problem.

[rex007can]

I am in the process of testing/formulating and disaster recovery procedure for our Scalix server.
So far it's a bit like this.
Install new RHEL3 and configure it.
Install Scalix 9.2 on it.
Test to make sure it works
Logged on to SAC
All is okay.

Stop scalix&tomcat
rename /var/opt/scalix
Restore /var/opt/scalix from backup (tar in this case)
Verify access rights.
Modify Authentication for local instead of AD
Restart Tomcat & Scalix
omcheck -i -d
omscan -a -A

It all seems to work. I am able to access my recovered mailbox through webmail using my original local password. (the same one I had to use for SAC with my user)

But...
I cannot logon to SAC.
In an actual DR case, I would reset all the user's local passwords and force change at next logon. I can probably do that through a script or at the command prompt. But logging on to SAC is somewhat important.

Is there something I'm missing for this part? Does the fact that the test machine does not have the same name and IP address have something to do with that? In actual circomstances, I'd configure the recovery machine the same as the original, but for obvious reasons I can't do that now...

Thanks.

[crazy_russian]

Hi,

A couple of questions:

Are you moving the message store from one machine to another machine that has a different hostname and IP address?

Is there a particular reason you are not upgrading to Scalix 10?

Thanks,
Yuri

[rex007can]

Yes, I guess in essense, I'm moving everything from one machine to the next.
The goal is to write a step by step procedure for this to put into our DR plan.

Yes, the "new" machine does not have the same IP and hostname as the "old" one.

And yes, I DO plan to upgrade to Version 10. As soon as this test is done, I run a test upgrade on it to make sure everything runs smoothly and to see where problems might arise, then I'll update the production server. Then I'll retest the DR procedure on Version 10. Should work pretty much the same.

I'm not comfortable doing a major version update on a production server before I've had a chance to see what will go wrong. Going from 9.2 to 9.4 caused bugs so from 9.4 to 10 surely will too.

[ScalixSupport]

We provide a utility to change relevant portions of Scalix when you change the hostname. It is called sxmodfqdn.

DESCRIPTION
sxmodfqdn makes the minimum necessary modifications to the Scalix server to ensure that, after the host name is changed, Admin Console will
continue to function. This involves changing the SYSTEM directory so that entries owned by this server will have the correct HOST-FQDN
attribute value, and changing the Admin Console sxqueryadmin user so that the authentication id will have the correct domain. Note that this
command does not make any change for other server settings that may be related to, or affected by, the host name - these settings must be manu-
ally checked and corrected individually.

This command should only be run on the Scalix server that has host name change, any other Scalix server within the network should be taken care
of by directory synchronization when the next update is due.


That should be incorporated as well.

Cheers,

Sascha.

[rex007can]

That did it!
Thanks a bunch!