This assumes that the ARCHIVE_FILE_CB1_NN constraint is lifted
before an import can be executed!

Giuseppe Lo Presti authored Jun 21, 2019 and Michael Davis committed Jun 24, 2019

CASTOR migration: alternate implementation to fill the directories intermediate table with no temp table

CASTOR migration: alternate implementation to fill the directories intermediate table with no temp table

The code now runs in parallel mode according to Oracle EM.

The full path is only stored in the CTADirsHelper table, whereas the
CTAFilesHelper table contains only the filename along with all other
relevant metadata fields.

Among others:
- need to create Logical_library entities if missing
- tapes in VMGR are marked as EXPORTED via PL/SQL

Following the current ATLAS CTA instance, the path prefix for CASTOR-imported files will be "/eos/cta<experiment>" (no "pps" and no extra slash)

CASTOR migration: several fixes in the PL/SQL code, which now compiles correctly on an "almost-final" CTA schema

Tape.media_type and Tape.capacity in CTA are uniquely
specified by the density in VMGR by a hard-coded mapping.
Also incorporated formatting changes, removed useless function,
converted file checksum to CHAR (this is temporary, pending
the final format to store file checksums).

Implemented latest decisions as documented in #488. The PL/SQL code is ready for testing once the schema changes are available.

Moreover, the PL/SQL code for the CTA catalogue is temporarily
moved out of the catalogue folder as it requires some schema
changes that are still work in progress in a dev branch. Until
then, all migration code is evolved in the migration/ folder.

References:
https://gitlab.cern.ch/cta/CTA/issues/479 (on the CTA side)
Jira: CASTOR-5526: support the migration to CTA (on the CASTOR side)

The latter won't be updated further, all required code will
be committed here and gitlab tickets will be updated accordingly.