If there is a Header entry for any given HD table with a very large amount of associated Detail lines, e.g. in excess of a hundred thousand or more, how well does NextWorld handle deletion of the detail lines when I delete that header entry? Taking into account this discussion, how does NW automatically perform efficient and guaranteed deletion of these detail lines?
Or in these scenarios, do we have to be aware that there could be a very large amount of detail entries and not rely on the cascading deletes (assuming this is how it’s implemented) and rather make use of a Data Purge?
The platform is built to process deletion of large amount of records, but that processing will be hindered based on the amount of validation work the system must perform to accomplish the delete. If your HD, Header, or Detail tables have triggers or there are Referential Integrity checks, performance can degrade significantly as the server validates the records one-by-one.
When a record is deleted via the UI, that process will complete as long as related validations do not fail, so a UI timeout will not stop the underlying operation. For larger scenarios, Data Purge is going to be your best bet, but there are situational reasons to use Logic Block actions instead.
Background Task logic (Delete Record logic actions): useful for controlling record deletion outside of the UI; useful for separating record deletions into smaller chunks that process in separate transactions. Not generally useful when a Header-Detail structure is at play.
Data Purge: this is built for high volume and programmatic record deletions. While it is more overhead and not useful for every situation, it will be faster.