documented what it means (Full documentation is anyway in DynamicThreadPool.hpp).

bug #89253: tapebridged sends end-of-session before the tapegatewayd has finished processing the session

Found a bug in the bug-fix.  I was trying to use body.rqst.diskFseq to access
a pending transfer in the following Indexed container:

  castor::tape::tapebridged::BridgeProtocolEngine::m_pendingTransferIds

The problem was I had, just a few lines before, removed the pending transfer
from the container.

bug #89253: tapebridged sends end-of-session before the tapegatewayd has finished processing the session

Improved the exception messages of the following class:

castor::tape::tapebridge::BridgeProtocolEngine

bug #89253: tapebridged sends end-of-session before the tapegatewayd has finished processing the session

Improved the exception messages generated by the methods of the following class:

castor::tape::utils::IndexedContainer

bug #89253: tapebridged sends end-of-session before the tapegatewayd has finished processing the session

First attempt (and hopefully the last) at fixing the out-of-order
end-of-session message.

This reverts commit e9748d9b94e170ad18bdca8d3db46b65a775e79d.

to drop Repack requests when they are just inserted without SubRequests
(Only case in the request handling logic). The solution is not perfect
though, there's a race between the cleaning job being too aggressive
and the NS answering too slowly for a Repack request to start getting
populated before the timeout. To be improved.

Reduced max number of concurrently starting/aborting repack to 3. A more conservative, thus safer number for a production environment

This is to fix the code base as the previous implementation was broken with the 2.1.12 schema,
pending discussion on the role of this procedure and whether the Lemon publishing upstream
should be fixed as well or completely dropped.

Fixed gathering of segments data from the nameserver. Note that the new code is ORACLE 11 specific. Thus repack will no support ORACLE 10 anymore

  - svcClass was dropped from the MigrationRouting table as unused. Tools have been amended accordingly.
  - MigrationJob has now an originalCopyNb and originalVID in case of repack migrations that allow to identify repacks properly
  - a table MigratedSegment has been added holding information on other existing tape segments for the migrating files. This allows to avoid putting 2 copies on a file on the same tape
  - in general, migrations that can be triggered at recall time (extra copy and repacks) are populated in MigrationJob table from the beginning, in status MIGRATIONJOB_WAITONRECALL. This allows easy triggering of the migrations after the recall and to identify any problematic migration routing before we even recall.

Fixed.