Rochdi, if you haven't created and practiced your own steps for doing journal restore, you might be best served by contacting the WRC or your account team at InterSystems.  Restoring from journals can involve many decisions about where the journals come from, where you are using them now, and several other items -- it can get complicated depending on exactly what you need to do.

Also, you can't restore a complete database from journals unless you have all the journals from the time you created the database, which isn't very likely if you have been using the database for more than a few days.  You also reported a very old version of Caché (2014.1) which might affect how you recover.

Certainly if you are in a crisis situation, contact the WRC for immediate help.

Alexander is observant.  There are five expansions, but these look like retries, because the size completed doesn't change.  Expansion is stuck for some reason.  The WRC might be able to find out more from ^SYSLOG if this is continuing to occur. 

Rochdi,
2.2TB is within the limits I could find for the NTFS file size limits and partition size limits.  Caché can grow a CACHE.DAT file to 32TB before reaching its software limit.

From what you've shared, I have a few ideas:

The NTFS partition is limited in size, causing the <FILEFULL>.

There is some Windows policy that is limiting the maximum file size to less than what NTFS allows.

You've encountered an incompatibility between Ensemble 2014.1 and Windows Server 2016 (which is not a supported platform for this version according to documentation -- it's too new).

Your best bet is probably contacting WRC for help in sorting this situation out.