Re: subtle performance question - simple query

[Sandro Frattura]

FYI. A response I got from a colleague on this (academic) question.

Hmm, I assume a trick question here --- but, based on what you provided, considering userFrom and userTo is a primary key, meaning you are querying against that specific, unique, row, the queries should be exactly the same, and one is not more efficient than the other.

Unless there’s an index on isRead and the optimizer is being completed HORRIBLE and using that index instead of the primary key. But that should not happen.

As far as when an update is happening or not, well, in the case where an update is not happening, it will still perform the action of looking it up, but it will realize that no changes need to happen so the data files need no change.