System Design — How to query status of 10,000 machines

The general answer is to 1) maintain a queue of 10,000 jobs, each one is to query a machine’s status 2) have a bunch of worker machines run in parallel, retrieving and executing one status querying job at a time.

Following are the steps of getting deeper in this question and ultimately reach a satisfying state.

Step 1 — Naive Approach

A pitfall when the interviewer presents this question is that, he won’t tell you how many machines you can use to do such an operation…