Scalix Forums

Hello,

I am in the process of upgrading my hardware , here is what I want to do.

I have a new system with better hardware than current system.
Install OS ( Redhat 4 )
Install 11.1 ( community )
rsync data from old server /var/opt/scalix
upgrade new server with 11.2
stop old server
Make the new server live

Will this process work, Please advice i fyou have a better process.

Sanjay

knowing what I know now, here is what I would do in your place:
1) Install OS ( Redhat 5, not 4 )
2) Install *ALL* Redhat 5 patches and upgrades.
3) Install 11.2 ( community, not 11.1)

Make sure everything works as you would expect for a fresh installation. This is to ensure all required subsystems to make scalix work are already installed. Never mind the IP or name, but at least have something properly configured to this"demo system" for DNS, because the install script will verify.

4) Use the install script to remove everything (yes, uninstall scalix)

Then more or less follow this steps on the post below:
viewtopic.php?t=8606#39043

I tried this but did not work , is there a way to just export and import users, groups and their settings. I can leave the mailstore on old server.

Can you produce some information on why you think it didn't work? This upgrade/migration process seems very streamlined within Redhat, and if for whatever reason somebody had a problem I think the whole community should know about it, and how to work it out. You won't be alone.

BTW, migrating the mailstore is the scalix way of exporting and importing users, groups and their settings.

Jaime,

I beleive it was because of the corrupt rsync transfer, my mail store size is 160GB and it takes about 10 Hours to complete the transfer.

After the rsync was complete I changed the name of the server to primary server, kept the IP same then I did a upgrade.
After the upgrade I could not logon to webmail or SAC it gave error user not present.

Do I need to change the server name and IP address on the new server ? to complete this .

I am trying to reduce the downtime as much as possible.

Reason for my upgrade is the IO bottlnect on the old server due to which local delivery was extremly slow some times the queue will increase to 5000.

Not all is lost.

1) rsync is a very nice command. If the bulk of the transfers is already done you can give further rsync's for to pickup the last changes, and subsequent cycles are much quicker. There is no such thing as a "corrupt" rsync transfer. What we have is improper syntax of the command or "live" changes on the original file system while the rsync is taking place. To address that just stop all services on the original server just before the last sync, including sendmail.

2) You **DO** need to change the FQDN and the IP on the new server to match the old one before the installation, and take the old server offline.

3) The thread below gives you a few more hints on possible problems, and the fixes:
- ensure you know how to use the -delete option on the rsync
- do a search on the forum on how to fix the permissions on the mailstore after the installation
- login problems can also be related to postgres. restart the DB and the server, and check if the services are running properly with "omstat -s and omstat -a"
viewtopic.php?t=8901