Aloha,

I am moving a long standing Xeams stand-alone server from a windows platform onto a Debian 8 machine.  It is a clean install.  I then follow the instructions to migrate.  I install Xeams on the new machine and all is working well.  I then move the data from the old machine except for the appconfig file.  (I have never located the wrapper file.) After moving, the program slows to a dead crawl and the only service I can get working is SMTP after binding to the new IP address.  IMAP and POP3 will not come to life.  The diagnostic check says everything is fine.

What am I missing?

Do you see any errors in Xeams.log file? It should be in /opt/Xeams/logs folder. Did you also move the /opt/Xeams/db folder? If this folder is missing or corrupted, Xeams will try to rebuild it. This folder holds an embedded database, which can take some time to rebuild.

The log file should have information if this DB is being rebuilt.

Yes, the log file shows repeated entries of "unable to get a database connection after waiting for 40 seconds."

I don't see anwhere about rebuilding though.  What can I do?  

It is very likely the database is corrupted.

Try the following:

• Stop Xeams
• Rename /opt/Xeams/db folder to something else so it is not found at runtime
• Restart Xeams

When this folder is not found, Xeams will try to recreate and re-populate data. It will run slowly for a few minutes. The actual time depends on how far your emails go in the past.

OK, but before I do that, I noticed it says  

SQLSTATE: XCL20Catalogs at version level 'null' cannot be upgraded to version level '10.3'

Should I deal with this upgrade issue or should I just rebuild the db?

Lastly, I had to put everything back on the original server again to stay up and running.  If I do this database repair, how can I migrate just todays activity without messing it up again?

Thank you

Don't worry about the version problem. Let it rebuild.

You will NOT be able to merge the two databases.  It will be difficult to move them to the older server if the newer server handles new messages.

The embedded database is only used to display messages without search. A corrupted database does not affect email flow. In the worst case scenario message won't show when trying to view them without searching, which will get fixed eventually since Xeams only displays messages for last X number of days.

There are two ways Xeams display messages through Message Repository:

• The embedded database - this is used when you try to show "View All Good/Possible/Junk". New entries will be inserted into the database as they newer emails come in. Records older than ten days (or whatever the retention period is) will get removed from the DB. 
• Search Indexes - these are used when a user explicitly searches for older messages. Database is not used in this case

Therefore, even if you can't get the DB migrated or recreated, the problem will go away in a few days, and it won't affect email flow.

So it sounds like the best approach is to NOT copy the db folder to the new server and delete the one that exists there already.  When I deleted it intially, it solved all problems in about 15 seconds.  

I'm going to have to do this migration again tonight and I'd like it to just go smoothly. This is the only way I can see to make that happen.