I am working on a Server which called a datasource parallely for different queries, takes up the data in memory, aggregates it and responds back. I have 4 cores. Till the time i was having 4 queries, my latency was normal, but since the time I have added 5th one, it's taking almost double the time. Also, the 5th Query is the quickest of all.
Ideally my CPU should not block on the IO of all the four to process the fifth one. But I wanted to understand when exactly does my CPU get blocked on IO.
- Does it get blocked when my server is receiving the response, but has not received it fully?
- Does it get blocked when my server has received the response and is parsing it? I believe it does.