I have a table that fills up with data very quickly. I have to read and analyze them quickly. (maximum 10 seconds).
To do this, I run 5-10 instances of my application.
I can't find the right approach: how to correctly read one table by several instances of the application so that each does not duplicate data?
I see it this way, every instance after reading puts a flag (read). But the rest of the instances will wait (if done with passive blocking)
How is this problem solved in large projects?
