I am currently using the Serverless Atlas deployment. The application i am trying to run on it, has many update operations on multiple collections. I am running a loadtest on my application and I am generating a constant load of 100-200 VU per second on the application and each thread perform 2 or 3 update operations. I am facing timeouts(as the database is taking a much longer time to perform designated task) after around 25-30 seconds the test starts, and then the timeout stops happening after around 240 seconds, same happens again after a considerable amount of load on the server. I am assuming that the database scales up, during that time which is around 3 minutes.
My question is, is that the standard scale up time required for serverless deployments? Should it happen faster, vis-à-vis is this a configuration malfunction?
Also, given a constant rate of traffic I am assuming on my application, does that mean i should look for dedicated deployments and not serverless?
Thanks in advance!