Reana-db

- Changes the `Workflow` table to replace the `run_number` column with two new columns `run_number_major` and `run_number_minor` in order to allow for more than nine restarts of user workflows.
- Changes the names of database table, column, index and key constraints in order to follow the SQLAlchemy upstream naming conventions everywhere.
- Changes several database index definitions in order to improve performance of most common database queries.

- Adds progress meter to the logs of the periodic quota updater.
- Changes CPU and disk quota calculations to improve the performance of periodic quota updater.
- Fixes the workflow priority calculation to avoid workflows stuck in the `queued` status when the number of allowed concurrent workflow is set to zero.

- Changes to PostgreSQL 12.13.
- Fixes conversion of possibly-negative resource usage values to human-readable formats.
- Fixes disk quota updater to prevent setting negative disk quota usage values.
- Fixes quota updater to reduce memory usage.

- Adds new `launcher_url` column to the `Workflow` table to store the remote origin of workflows submitted via the Launch-on-REANA functionality.
- Adds the possibility to force resource quota updates irrespective of globally-configured quota update policy.
- Adds new `WorkspaceRetentionRule` table to store workspace file retention rules.
- Adds new `WorkspaceRetentionAuditLog` table to store the audit log of workspace file retention rule updates.
- Changes percentage ranges used to calculate the health status of user resource quota usage.
- Changes to PostgreSQL 12.10.
- Fixes wrong numbering of restarted workflows by limiting the number of times a workflow can be restarted to nine.
- Fixes `Workflow.get_workspace_disk_usage` to always calculate disk usage rather than relying on the quota usage values from the database, since these may not be up-to-date depending on the global quota update policy.
- Fixes helper function that retrieves workflows by UUID to also additionally check that the provided user is the owner of the workflow.

- Adds transition for workflow from queued to failed status.

- Adds an option to periodically calculate CPU quota usage.
- Changes CLI quota command from `disk-usage-update` to `resource-usage-update` since it can also perform CPU quota calculation.
- Fixes quota update functions to handle exceptional situation as continuable errors.
- Removes extra `QuotaResourceType` enum in favor of `ResourceType.name`.