Scalix Forums

Hi,
I tried to upgrade to 11.0.1. So I extracted the tarball from the old suse 10.0 server to the new one (10.1). Then I created a user "scalix" with the uid 103 (the same as at the old machine).
The next step was to install scalix 11.0.1. The installation proccess seemed to went fine.
Login to sac works.
Login to webmail shows me the "evaluation" error message.
A restart of tomcat (/etc/init.d/scalix-tomcat restart) does not resolve the problem.

omcheck -i -s shows no need of modification.

omstat -a shows:


A difference between old and new server: they have different IP addresses.
/etc/hosts seems to be ok:


omshowlog shows only some ERRORS and WARININGS:



Do you have an idea, what I did wrong?

Some new info:
I changed the permission to group scalix (chown scalix:scalix /var/opt/scalix -R)

Now I can log in with some users. (some login fails with username or password incorrect)

Now omshowlog shows many errors of the types:

Hi!

Can you check the permissions for scalix files using the command "omcheck -i -s"
To actually fix the problem if any of permissions, you should run omcheck to generate
a shell script - on standard output - that, when run, fixes the problem for permission
issues.

You might want to review what the script is planning to do before actually executing the second command.

Thanks,
Subir

Hi,
I have checked with omcheck, but all seems fine concerning permissions.

I have already tried the upgrade from start, but I still have the same fault.

Is there something else, I can do?

Bye,
NP

Hi!

Did you run the shell script generated by the command I suggested in my last post?
If you ran the script and still face the problem, I would suggest you to run the command
below for those users for whom login fails:

This pre-generates the mime data for the messages and should resolve the problem.

If the above step does not help either, then I would request you to perform the steps
below to try and reslove the issue:

You need to watch the status of the omscan process using the command:

Wait until you get a message that says "Current server cycle not started;
service reset or delayed".

Now try to fix the mailstore related issues in active mode using the below given
command:


Thanks,
Subir

As the file ownership under ~/user or ~/data seem to be affected, you will need to run omcheck as



Florian.

I hit this last night migrating from SLES to RHEL for a client.

omcheck is missing some directories. It SAYS that all is fine, but it isn't. I'll enter a bug into bugzilla shortly.

Kev.

Hi,
so there is my next question: ;)
Should I try to solve my problem by executing the hints above or do I have to wait, until a fix is released?

Bye,
NP

Well, running omcheck with all possible options will certainly do no harm.

If there is anything omcheck is missing, that would need to be logged as a bug and we can look into it.

However, as there is a very clean "workaround" (which is actually more like a best practice), i.e. to copy the files with a method preserving ownership and permission in the first place, we're not likely to treat this with a very high priority.

Cheers,
Florian.

So is there a difference to "my" way?
I thought, make a tarball with parameter p, copy it to the new server and extract should be enough ...
Do I have to add the user and group "scalix" before or after the extracting of the tarball?
And the UIDs should be the same as at the old machine?


Bye,
NP

The issue I ran into was that the Scalix user had a different ID on the two boxes. Likely due to a distro change.

Kev.

This can easily be helped by copying over the files using tar. Gnu tar will by default rather go with the symbolic username and adjust file ownership accordingly if he user has a different UID.

Being a Unix guy of 20 years experience with various -x-type operating system that surprised me quite a bit when I learned it (as tar really had never been that way), but i've learned to accept the convenience of it.

So, as long as you create the scalix user and group before unpacking the archive, you're safe when using tar, which is what I would personally always recommend for the very reason that it seems to be the most permission save archiving tool.

Florian.

Good to know.

I used rsync so I could get 99% of it while the server was up and running, and then after I shut the server down, it was just a 5 minute resync to make it current. The initial sync took 12+ hours.

Oh well. I'll know for next time...

Kev.

BTW, my omcheck issues was a problem between the keyboard and the chair I was sitting in... :)

Kev.