Design Patterns for Scaling - The Practice of Cloud System Administration - page 116

Exercises
1. What is scaling?
2. What are the options for scaling a service that is CPU bound?
3. What are the options for scaling a service whose storage requirements are grow-
ing?
4. The data in Figure 1.10 is outdated because hardware tends to get less expensive
every year. Update the chart for the current year. Which items changed the least?
Which changed the most?
5. Rewrite the data in Figure 1.10 in terms of proportion. If reading from main
memory took 1 second, how long would the other operations take? For extra cred-
it, draw your answer to resemble a calendar or the solar system.
6. Take the data table in Figure 1.10 and add a column that identifies the cost of each
item. Scale the costs to the same unit—for example, the cost of 1 terabyte of RAM,
1 terabyte of disk, and 1 terabyte of L1 cache. Add another column that shows the
ratio of performance to cost.
7. What is the theoretical model that describes the different kinds of scaling tech-
niques?
8. How do you know when scaling is needed?
9. What are the most common scaling techniques and how do they work? When are
they most appropriate to use?
10. Which scaling techniques also improve resiliency?
11. Describe how your environment uses a CDN or research how it could be used.
12. Research Amdahl's Law and explain how it relates to the AKF Scaling Cube.