The gRPC protobuf will be shared by EOS, CTA and CERNBox, so it needs to
go in its own submodule.

Added the ArchiveForRepack mountType to the Drive reporting
Transformed the DataTransferSession::mountTypeToString() to add the ArchiveForRepack mountType, transformed the normal "Archive" into "ArchiveForUser"

Update cta::Catalogue::tapeMountedForRetrieve and
cta::Catalogue::tapeMountedForArchive to increase read/write
mount counts.

Added some unit tests to test all garbage collection of all type of Retrieve Requests (ToTransferForUser, ToReportToUser, ToReportToRepackForSuccess, ToReportToRepackForFailure)

Honoured the individual Retrieve garbage collection for jobs ToReportToRepackForSuccess and ToReportToRepackForFailure

We populate only the CHECKSUM_ADLER32 column (NUMBER) during migration.
CHECKSUM_BLOB is filled for new files not migrated files.

containing: eos-import-dirs

And renames tool to cta-upgrade-db in line with other tools

Added XRDSITE=cern_tape_archive_ci to eos sysconfig file in order to tell clients that there is a CTA backend

Commenting this out from master due to broken dependency in CI

- CASTOR migration tools stay under CTA/migration
- upgrade tools are now under CTA/upgrade_db
- adds FindGRPC.cmake (required for migration tools)
- updates submodule to v0.50 (required for EOS gRPC API)

rdbms::Rset::next() now releases the underlying resources of a result set when there are no more rows to be fetched.

Added much needed comment to the setting of the CLIENT_FOUND_ROWS flag when making a MySQL database connection.

Add isFromCastor field to the catalogue DB.

CTA only sets this filed to false and respects it in the getTapesForWriting

Changed 'xrootd.fslib libXrdEosMgm.so' to 'xrootd.fslib -2 libXrdEosMgm.so' so that XrdSfsGetFileSystem2() is called which in turn tells XRootD that an alternative Prepare plugin is present.

Sets the ChecksumBlob from the supplied protocol buffer, unless the
protobuf is invalid, in which case it uses the supplied Adler32 checksum
instead.