Tips suggested by Oliver Shulz

First, the problem with negative OB IDs (which cannot be logged into the NightLog) can have several origins, including - among
others:

- an incorrect IP in the instmgr@wxxxdhs account (e.g., an
   incorrect IP version number or compilation errors might cause
   this problem already)
- corrupted or lost local p2pp configuration files in
   visitor@wxxxdhs

At the 2.2m there is a further problem that some VAs (for special
reasons) used to run p2pp under the service account instead of visitor@w2p2dhs. 
Use of p2pp under visitor and service account in parallel can have unpredictable results.

Possible solutions:

- Ensure that only one p2pp version is running (visitor account).

- Re-init of some p2pp config files by starting p2pp on
   instmgr (!) account at w2p2dhs. Make a FULL log-in with a valid
   account. Then return to the visitor account.

- Somebody may need to look at the 'instruments' subdirectory on
   instmgr@w2p2dhs whether there are any IP inconsistencies.

Interestingly, some occurrences of LPRS-010212 may point to an eventual mismatch of parallel p2pp sessions.

Second, it is definitely not normal that for VISITOR runs the clone-to-database is not active by default. 
For all visitor runs, this icon should be ticked by default and grey-shaded (i.e. not selectable). 
Only TEC accounts should have the option to activate or deactivate clone-to-database.

If no problem on instmgr@w2p2dhs is found, then there might eventually be a database problem, i.e. 
some VA accounts might be flagged in the database as TEC accounts (you can ask DHA about this).

Let me know if you need more details. However, without being able to see details of the problem, 
my advise is unfortunately limited to experiences from earlier occurrences of similar problems.