Part 2: Build M/M/m model

[taivop]

Complete by Thu 08 Dec 2016!

TODO:

• Think how to calculate arrival rate and service rate for the model (service rate: max over all experiments).
• Think how to handle repetitions: SOLUTION just consider one rep from each #clients.
• Wrap calculations in functions that take experiment folder as a parameter, so I can iterate over them.
• Use exp1 (throughput) results and vary num of clients (threads constant at 32).

---

Depends on #170.

• entire system
• multiple M/M/m models; compare to multiple experiments
  • don’t need to do for all experiments of 2nd milestone, just a few with different configurations.
  • vary # clients or # servers or replication
  • take scalability graphs (response time as function of S)!
  • compare in unsaturated and saturated regions
• more than 1 model for more than 1 configuration