Non-functional System Characteristics

Question 1

Name all the non-functional requirements

Answer 1

1. Availability
2. Reliability
3. Scalability
4. Maintainability
5. Fault tolerance

Question 2

Define availability

Answer 2

% of time a service or infrastructure is accessible to clients and operating under normal conditions

Question 3

Availability equation

Answer 3

A = (total time - downtime) / total time

Question 4

Does availability include planned down time?

Answer 4

No

Question 5

Describe amount of downtime for 1-9, 2-9, 3-9, 7-9s

Answer 5

36.5 days, 3.65 days, 9h, 3s

Question 6

Define reliability

Answer 6

Describes how a service performs until varying conditions. Probability that a service will perform its functions for specified time.

Question 7

What are the two reliability metrics and their equations?

Answer 7

1. MTBF - mean time between failures
   - (total elapsed time - downtime) / # failures
   - higher is better
2. MTTR - mean time to repair
   - total maintenance time/# repairs
   - lower is better

Question 8

What constitutes SLA/ SLO?

Answer 8

Availability (time lost)
Reliability (freq and impact of failures)

Question 9

Is reliability a function if availability?

Answer 9

No

Question 10

Is availability a function of reliability?

Answer 10

Yes

Question 11

Describe the two ways of scaling in scalability.

Answer 11

1. Vertical scaling or scaling up
   - add extra capabilities to existing device for single huge server
2. Horizontal scaling or scaling out
   - increase number of machines in the network in a distributed manner
   - system now works with many nodes

Question 12

What are the three requirements of maintainability?

Answer 12

1. Operability - ease with which we can ensure smooth ops under normal conditions and normal ops during failures
2. Lucidity - simplicity of code
3. Modifiability - ability to integrate modified or new features

Question 13

Define maintainability and it’s metrics

Answer 13

Probability system will restore functions within a specified time of fault occurrence. Use MTTR, as low as possible

Question 14

What’s the difference between maintainability and reliability?

Answer 14

Reliability = MTTR + MTBF
Maintainability = MTTR only

Question 15

Define fault tolerance

Answer 15

Systems ability to execute persistently even if one or more components (HW/SW) fails. Need to avoid single point of failures by ensuring data safety and avoiding expensive recalculations.

Question 16

What are the two common ways of ensuring fault tolerance?

Answer 16

1. Replication: replicate services and data
   - sync: A low but consistency high
   -async: A high but eventual consistency
   - CAP theorem
2. Checkpointing: save system state in stable storage when system is consistent
   - has same problem as replication