System Design interview
I was giving a System Design interview where the problem was to design a metrics system.
The interviewer asked: What would you do if the database is almost full and you are not allowed to scale it? How would you store new incoming metrics data?
I suggested one approach - aggregating old data already present in the database , but the interviewer was also expecting other possible solutions.
Can someone help me with this problem? What are the different approaches we can use to handle new metrics data in a constrained storage environment?
(caution: have not worked with large scale systems ever, just know some theory)
What does "not allowed to scale" mean in this context?
Does that include sharding/partitioning? What about storing in column major, i.e. inverted rows and columns? Or Archiving based on usage patterns? Dead letter queue until a new db is provisioned?
