CHAPTER 6 · 2 lessons · 26 exercises
System design and performance
How a system is built so it does not go down, and how speed gets measured.
This chapter is about what to do when one server is no longer enough, and about keeping things running when something breaks.
Client and server and how the work is split between them, redundancy (duplexing and clusters), calculating availability, response time and throughput, and the cloud service models (IaaS, PaaS, SaaS). Plenty of questions here want a number out of you, but there are only a few formulas to use.
Lessons in this chapter
Configuration and redundancy
- 13Building so a breakage does not stop youWays of splitting the work across several machines, and what you keep ready so one failing machine does not stop everything. Duplication, RAID and the cloud.Go to the exercises
Uptime and performance
- 14Counting how little it stops and how fast it isHow much does that system stay up, and how fast is it? The handful of formulas and ways of counting that let you say it with numbers rather than impressions.Go to the exercises