Design a Metrics Alerting Engine
This prompt comes from Palo Alto Networks software engineer loops between January and March 2026, where one candidate got it as the design round of an onsite. The engine also turns up on the coding side. Another candidate's phone screen was built around alert engine milestones. An onsite coding question asks for the core loop itself. Load the alert configs, run each query on its interval and compare to a threshold, then notify while critical and send a resolve when the value drops back. So you may be asked how that loop works in code, and then how it holds up at 100,000 rules.
The metrics database already exists. Firewalls, agents and services write samples into it, and you can query it but you don't own it. Your system sits on the read side. It decides when each rule runs, asks the database a question, remembers the answer, and tells a person when something changes. That makes this a scheduling and state problem more than a storage problem. The data you keep is small. What's hard is running a hundred thousand small jobs on time, forever, while machines fail and users edit rules underneath you.
The guide's scalable alert manager design takes alerts that sensors already raised and dedupes, groups and routes them. This prompt is the part that design leaves out: deciding that a metric is in trouble in the first place. The two fit together, and the notifier here can hand its output to that pipeline.
Palo Alto NetworksPremium lesson · System design round
The Palo Alto Networks design, worked end to end.
- Requirements and scopeSystem design
- The API and data model
- The design, drawn out1 diagram
- 5 deep divesWith the tradeoffs
$49 a year · all 107 companies · instant access
20,000+ candidate reports · guide updated
Palo Alto Networks interview guideKnow what Palo Alto Networks asks before you walk in.
28 coding · 4 system design · 15 behavioral questions
Get their questions first.
InterviewLegend is independent and not affiliated with or endorsed by any company it covers.