Skip to main content
TT
Lab
Learn
Sandbox
Community
Language Learning (in Korean)
Get started
Learn
Learning paths
Courses
Observability
Run PromQL, SLOs and alerts on a real Prometheus
중급 · Lessons 30 · Lab 10
Start the lab
Curriculum
Metrics, Logs and Traces
The Three Signals Are an Investigation Order, Not Substitutes
reading
Throwing Queries at a Real Prometheus
lab
Quiz: The Three Signals and Metric Design
quiz
SLOs and Error Budgets
The Error Budget Is a Negotiating Tool
reading
Computing the Error Budget as a Number
lab
Quiz: SLOs and Error Budgets
quiz
Alert Design
Some Alerts Are Accurate and Still Harmful
reading
Pushing Alerts Through Alertmanager
lab
Quiz: Alert Design
quiz
Cardinality Explosion
Why You Must Not Put user_id in a Label
reading
One More Label and the Series Count Went Up a Hundredfold
lab
Quiz: Cardinality Explosion
quiz
What Exactly Are You Counting?
Change the Denominator and Availability Changes
reading
Same 12 Hours, Availability of 98.9% and 85.0%
lab
Quiz: Defining an SLI
quiz
Alert Fatigue and Symptom-Based Alerting
Enough Accurate Alerts Add Up to an Inaccurate One
reading
Seventy Alerts Overnight and Users Noticed Nothing
lab
Quiz: Alert Fatigue
quiz
Where Graphs Mislead
A graph is not the data, it is a summary of the data
reading
The same outage was invisible on the dashboard
lab
Quiz: Reading Graphs
quiz
Log Levels and Sampling
Keep every log line and you will find nothing
reading
The day we turned on debug logs, the lines we needed disappeared
lab
Quiz: Log Levels and Sampling
quiz
Finding the slow part with traces
The longest span is not the culprit
reading
We fixed the slowest span and the response time did not move
lab
Quiz: the critical path
quiz
Incident investigation and postmortem
The first thing an investigation does is pin down the time
reading
Two of the three hypotheses could have been ruled out from data in 30 minutes
lab
Quiz: incident investigation
quiz
Reference docs
Prometheus 문서
Grafana 문서
Google SRE — 모니터링