DEVSPS-SRE practice question 9 of 10

Which of the following BEST defines the golden signal for errors?

Full set: 80 questions
Answer options
A: The time it takes to service successful as well as failed requests
B: The rate of failed requests either explicitly implicitly, or by policy Correct answer
C: The demand that is placed on your system by me volume of requests
D: The percent of capacity used by your system for current requests
Correct Answer:
B. The rate of failed requests either explicitly implicitly, or by policy
Explanation

The "four golden signals" are a set of key metrics recommended for monitoring a user-facing system. The 'Errors' signal specifically tracks the rate of requests that are failing. This failure can be explicit, such as an HTTP 500 error code, or implicit, where a request returns a success code but the content is incorrect or incomplete. It can also be defined by policy, such as a request that exceeds a latency threshold being counted as an error. Monitoring this signal is critical for understanding the immediate health and correctness of a service.

Why Incorrect

A. This describes the golden signal of Latency, which measures the time it takes to service a request.
C. This describes the golden signal of Traffic, which measures the overall demand on the system.
D. This describes the golden signal of Saturation, which measures how "full" the service is in terms of its constrained resources.

References

1. Beyer, B., Jones, C., Petoff, J., & Murphy, N. R. (2016). Site Reliability Engineering: How Google Runs Production Systems. O'Reilly Media. Chapter 6, "Monitoring Distributed Systems," defines Errors as "The rate of requests that fail, either explicitly (e.g., HTTP 500s), implicitly..., or by policy."

2. Schulman, A. (2018). Monitoring and Alerting for SREs. In Proceedings of the 2018 USENIX Conference on System Administration (LISA18). Section 2.1, "The Four Golden Signals," reiterates the Google SRE book's definition for the error signal.

Scroll to Top