[Improvement]: Improve processing of generic data index queue

[mcop1]

Changes in this pull request

Resolves #337
Resolves https://github.com/pimcore/service-operations/issues/173
Resolves #343

Additional info

Improvements adapted from #337

• Added id column with auto_increment as pk
• Added unique index for elementType and elementId (former pk)
• IDs will be added to the queue in batches now. Each batch/chunk will have its own transaction, therefore reducing the time a lock is held, if a lock was required
• Optimized id retrieval from queue using non blocking ids only (FOR UPDATE SKIP LOCKED)

Additional improvements

• Adding Update/add operations to the queue, for DataObjects, Assets and Documents, are handled async now in most of the cases.
• Added type filter when selecting related ids for data objects. This will allow better use of data base indexes.
• Added check for affectedRows: Updating the chunk will happen only if the insert returned affected rows lower than the actual chunk size.
• Refactored delete statement to include the id column, order by and limit.

Note: Delete operation isnt async for now, since we cant access the deleted element in the backend anymore. To achieve this, more refactoring is needed.

[herbertroth]

@mcop1 Please make sure the required reviewer is assigned.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
7 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud