[PGSCHED] design scheduler DB structure for archive job queue

Currently all info from the objectstore related to any job has been put into one row of the new database schema.

I suggest to go through all the columns to understand more globally:

1. origin of the information per column (clarifies the worlflow)
2. availability when filled/fill order (helps prevent updates [1][2] and rather do inserts when needed)
3. which filter fields we use to filter rows systematically for selection ? (helps understand which fields should be indexed)
4. which we systematically select for full job execution after insertion ?
5. systematic update ? (helps to recognise where updates are really necessary and where we can insert a new row instead [1][2])
6. mountID vs TapePool concepts could direct the Archive and Retrieve jobs in separate tables (execution related tables rather than queue related)
7. we design a queuing system - throughput is essential !
8. anything else you might think of ?

These should help us design a Archive related table(s) in the new PGSCHED DB schema.

The current tables have the following form (will try to make mermaid diagrams later to make the picture clearer):

TABLE:
REPACK		repack_reqid	status	create_time	mount_policy	vid																								is_complete	is_no_recall	is_move	is_expand_started	is_expand_finished	is_add_copies	create_username	create_host	user_provided_files	user_provided_bytes	total_files_to_retrieve	total_bytes_to_retrieve	total_files_to_archive	total_bytes_to_archive	retrieved_files	retrieved_bytes	archived_files	archived_bytes	repack_finished_time	last_expanded_fseq	buffer_url	failed_to_create_archive_req	failed_to_archive_files	failed_to_archive_bytes	failed_to_retrieve_files	failed_to_retrieve_bytes	subreq_pb	destinfo_pb
RETRIEVE	job_id	retrieve_reqid	status	creation_time	mount_policy	vid	mount_id	starttime	priority	storage_class	retrieve_min_req_age	active_copy_nb	size_in_bytes	archive_file_id	checksumblob	requester_name	requester_group	dst_url		disk_instance	disk_file_path	disk_file_id	disk_file_gid	disk_file_owner_uid	disk_system_name	repack_reqid	is_repack	retrieve_error_report_url	failure_report_url	failure_report_log	activity	is_failed	is_verify	srr_activity	srr_mount_policy	srr_time	srr_host	srr_username	lt_create	lt_first_selected	lt_completed	rr_repack	retrievejob_pb	info_pb
ARCHIVE	job_id	archive_reqid	status	creation_time	mount_policy	vid	mount_id	start_time	priority	storage_class	min_archive_request_age	copy_nb	size_in_bytes	archive_file_id	checksumblob	requester_name	requester_group		src_url	disk_instance	disk_file_path	disk_file_id	disk_file_gid	disk_file_owner_uid		repack_reqid	is_repack	archive_error_report_url	archive_report_url	failure_report_log	failure_log	repack_dest_vid	is_reportdecided	total_retries	max_total_retries	retries_within_mount	max_retries_within_mount	last_mount_with_failure	total_report_retries	max_report_retries	tape_pool	repack_filebuf_url	repack_fseq

[1] https://malisper.me/postgres-heap-only-tuples/

[2] https://www.postgresql.org/docs/current/storage-hot.html

Related also to issue #604 (closed).