I have a large dataset
Table: id | info1 | info2 | ...
There are multiples process accessing the data heavily.
Is there a built-in way in postgresql or an extension to know the number of time a row is accessed with SELECT ?
I was thinking of making a stored procedure and manually manage it.
No, there is none in any DB I know. It would be hard to implement too. Sometimes you don't need to access physically the data on the drive because you either find it in cache or take it from the index. Such procedure - if existed and worked according to requirement - would hit the performance of the DB hard.