A visitor says the page feels slow. Start with what the visitor experienced, then choose a measurement that can describe it. A signal is an observation such as a count, a measurement, or a saved event.
Service Monitoring · Lesson 1
Monitoring Signals
Learn to choose a monitoring signal that answers a question about a web service.
Checking the User's Experience
Availability describes whether the service can do the expected work. A check might request the homepage and verify a successful response. Latency is the time the request takes. A page can open successfully and still take several seconds.
Visitors receive the page, but it takes several seconds. Which signal describes the symptom most directly?
Counting Requests and Errors
Traffic is the demand on the service, such as requests per minute. Track the total alongside the count of failed requests. Five errors during 1,000 requests tell a different story from five errors during ten requests.
You see five failures in each of two windows. What helps compare the user impact?
Checking for Resource Pressure
Saturation tells you how close a resource is to its capacity. A service might have work waiting for its database even when the web process is running. Compare user-visible behavior with the resource signal before drawing a conclusion.
Requests slow down at the same time as the backlog grows. What is a reasonable next step?
Lesson complete
You finished Monitoring Signals
You can now choose a signal based on what a user noticed.
Distinguish availability and latency.
Interpret failures alongside request volume.
Use saturation as a clue about capacity.
Keep your learning progress
Create a free account to save this lesson and continue learning on any device.
Create a free account