Question

After performing an insert/update/delete, is it necessary to query the database to check if the action was performed correctly?

Edit:

I accepted an answer and would like to use it to convince management.

However, the management insists that there is a possibility that an insert/update/delete request could be corrupted in transmission (but wouldn’t the network checksum fail?), and that I’m supposed to check if each transaction was performed correctly. Perhaps they’re hinging on the fact that the checksum of a damaged packet can collide with the original packet’s checksum. I think they’re stretching it too far, and in most likelihood wouldn’t do it for my own projects. Nonetheless, I am just a junior programmer and have no say.

Answer 1

Shouldn’t be. Commercial database inserts/updates/deletes (and all db transactions) follow the ACID principle.

Wiki Quote:

In computer science, ACID (atomicity,
consistency, isolation, durability) is
a set of properties that guarantee
database transactions are processed
reliably.

If you have the feeling that you need to double check the success of your transactions then the problem most likely lies elsewhere in your architecture.