- avoid looping too much in the tapegateway when some file is locked for long
  - improved handling of log messages when the DLFLogs table is too full :
    + do not take all of it in one go
    + use an index as the table may be scattered

I have made not being able to execute the no-more-work flush to tape a fatal
error for rtcpd.

The refactoring carried out for this RFE has uncovered what maybe a potential
weakness in the code used to flush data to tape.  This modification is based on
what we have learnt about the legacy code during its refactoring.  It improves
the safety of the tape flush logic.  The current code in production has never
failed, however it is always better to be safer than sorry.

Fixed in the trunk.

bug #102830: RFE: Remove unused accounting from the tape server software
bug #102831: RFE: Remove the reason argument from tpconfig

I have put back the following line because the building of the CATSOR rpms fails
without it.  I am nonetheless confused as to why we need it in the tape server
makefiles seeing as neither the taped or rtcpd daemons generate system accouting
messages:

LOGROTATE(castor-sacct,installtape)

Removed the rtpcd accounting code.

Removed more tape-server accounting-code (hopefully the last).

Fixed in the master branch.

Removed the accttape C structure and the tapeacct() function protoype
which is no longer used.

*Fixed in the master branch*

        This is just for maintaining compatibility with the current deployment

upgrade, and modified to work around an Oracle 'feature' that
makes the procedure slower and slower over time.

Added more dependencies: all internal dependencies are qualified with version-release, and oracle dependencies are explicit

Updated the ReleaseNotes by adding a skeleton layout for 2.1.14-3 that includes the current tape bugs and features

Trying the client build in TeamCity without "XBS-Group: Client".

Fixed in the master branch.

Fixed in the master branch.

Fixed in the master branch.

before the spelling was correct but the capitalization was wrong.

Fixed the fix.  Thanks to me (Steve) we were trying to print an integer as a
string.  This caused a segmentation fault in taped.

Fixed in the master branch.

used to mount, dismount and query tape volumes on ACS-compatible tape-libraries.

- Added more details about castor.conf changes
- Reverted back to a generic phrase the explanation about
  the post-upgrade phase for the Nameserver DB.