Scalix Forums

Hello,

I am trying to move my scalix server to another machine. Currently I am running Scalix 11.1 on openSUSE 10.1 and I plan to move to Scalix 11.2 on OSS 10.2. If this works I will afterwards try to move to a production server like centOS. I searched through the forums and I tried several times to do the move but it always fails when I try to start the scalix services.

The steps I followed for the move I got them from one of Florian's posts:

1. Build the new system preserving hostname and IP
2. Make sure you create user and group scalix with the SAME Linux user and group id as on the old system (copy entries in /etc/passwd, /etc/shadow, /etc/group)
3. Copy /etc/opt/scalix/instance.cfg and the whole /var/opt/scalix/??/s directory. I used tar to archive the directory, moved it to the new server and extract it
4. Run Scalix Installer. It should see/find the existing message store and pick it up.

now, when I type /etc/init.d/scalix start it says FAILED

I am stuck. The installer log says that install was successful. Can anyone point me in the right direction.

thx in advance for all the help

Can you post the output from /var/opt/scalix/??/s/logs/fatal.

Also the output from omstat -s and omstat -a

Thanks
Kev.

Kev,

there is nothing in fatal log and there are no other log files created in this folder. When I moved the data store I deleted the old logs thinking that the errors will be written in new log files. I need to start figuring where to start looking for the problem.

omstat -a


omstat -s

thx
~mircea

so if you type omrc, none of the logs (especially fatal) are recreated, and there's no useful errors given?

My gut is that this will be permissions. Can you post the output from "ls -alh /var/opt/scalix/??/s/" I don't know it for sure, but that's my immediate gut feel...

Kev.

Thanx Kev for the fast reply.

I run the omscan tool and found some file permissions that were improper. Run the script and fixed those. Next I run the installer again to reconfigure scalix components but still nothing started.

Next step, I uninstalled scalix completely keeping the message store in place. I reinstalled scalix and it still does not start. I do have a fatal log now that shows this:



when i type omrc i get this error message


I know I have seen this error before in the forums but I cannot find a solution for this

this is the listing of the s directory


thx again for all your help

~mircea

Kev,

I managed to install the server. I found the problem to be connected to the different version that the message store had in /var/opt/scalix/fa/s/sys/version

I changed the entry in this file to the one from omvers and then reconfigured the server using the scalix installer.

this post was useful in that direction http://www.scalix.com/forums/viewtopic.php?t=6920&postdays=0&postorder=asc&start=15

thanx for your help

best regards

~mircea

Hello,

we are just at the point where we will do the very same thing.
Are the steps below still the "correct" ones or did you have to do something additionally ?

mdraghici wrote:Hello,
1. Build the new system preserving hostname and IP
2. Make sure you create user and group scalix with the SAME Linux user and group id as on the old system (copy entries in /etc/passwd, /etc/shadow, /etc/group)
3. Copy /etc/opt/scalix/instance.cfg and the whole /var/opt/scalix/??/s directory. I used tar to archive the directory, moved it to the new server and extract it
4. Run Scalix Installer. It should see/find the existing message store and pick it up.

In step 4. did you install Scalix 11.1 or directly 11.2 ?

André

That error, about the different versions of libraries and executables should be resolved by running ompatchom.

Even though the versions match now, I'd recommend running it anyway.

It'll take a couple of mins to complete.

Kev.

Andre, in step 4 I installed directly 11.2. Kev, thanx for your help. I will run ompatchom to make sure everything works properly

Hello,

last night we did sucessfully upgrade from Suse 10.1 Scalix 11.1.x to CentOS 5 and Scalix 11.2.

Of course there where some obstacles, otherwise I would have got too much sleep

Basically I did it as was mentioned in this thread.
The obstacles:

- The datapartition was reiserfs, and Centos default kernel does not come with reiserfs support. After adding the centos plus repository and installing the new kernel as described here http://wiki.centos.org/Repositories/CentOSPlus I had access to the data again.

- During the install I got the error message that it could not start postgres. It turned out that I had to remove all /var/opt/scalix/??/* files, except the /var/opt/scalix/??/s folder. (I did a inplace upgrade)

- The file /var/opt/scalix/??/s/sys/version did not got updated with the new version, so I had to update it manually

- Last the apache did not start because the file /opt/scalix/global/httpd/scalix-web-client.conf was missing. I copied it over from the old installation and then things got running. This one is a bit strange, but might result from the fact that I had removed the 11.2 packages and reinstalled them once more.

But now everything is OK and running smooth.
The installer did a good job, thanks to you all.

André