Optimize StrawberryFlavor (OCR, etc) Indexing/Updates and aggregation

[DiegoPino]

What?

For the sake of consistency we do a lot of update/tracking/parent and child re-tracking on any changes on ADOs and Flavor Data Sources. But on a large production server a few ADO updates that really don't affect at all Flavors end reindexing 10K++ (add an extra +) Flavor Documents and the number of Queries/Back and forth that are happening are/might end being too much with danger of timeouts and server slowdown.

We can Optimize and make better decisions on when these heavy Server tasks are needed using more logic/conditionals and maybe even aggregation tasks into single backend queue items to avoid so much processing.

Pieces we need to touch are:

https://github.com/esmero/strawberryfield/blob/1.1.0/src/Plugin/search_api/processor/StrawberryFlavorAggregate.php

https://github.com/esmero/strawberryfield/blob/1.1.0/src/EventSubscriber/StrawberryEventSaveFlavorSubscriber.php

https://github.com/esmero/strawberryfield/blob/1.1.0/src/Plugin/search_api/datasource/StrawberryfieldFlavorDatasource.php

And their interactions. There is a lot of Dinosaur-egg-Dinosaur situations in these parent-child interactions and sometimes the lack of context (is the operation an Update or a First Time Index?) makes all more complex

@alliomeria @aksm @karomabiles This is what was discussed today

Thanks