# Design an uptime monitoring service: system design interview question

> Regional probe results can disagree, so the service must distinguish target outages from probe location failures.

Source: https://sysround.com/questions/design-uptime-monitoring-service
Updated: 2026-11-05


Difficulty: hard. Levels: Senior, Staff+.

## How the interviewer opens it

"Design a service that checks websites and APIs from several geographic regions. It must report outages quickly without false alarms from a single probe location."

## What to clarify first

- Monitor definition
- Scale and frequency
- Read and write traffic
- Probe outcome
- Outage and recovery rules
- Missing evidence
- Latency
- Consistency and durability
- History and notifications
- Scope and security

## Scope of a strong answer

Core requirements:

- Create HTTP monitors with three probe regions.
- Run probes on schedule and retain recent results.
- Confirm outages and recoveries without false alarms from one probe location.
- Show current status and send incident webhooks.

Usually out of scope:

- Browser scripts and custom probe code.
- Private network targets.
- Billing and account management.
- Changes to the target website or API.

## What interviewers probe

### Outage decisions

A failed request can indicate a target outage, a probe location failure, or missing evidence. These cases require different status decisions.

### Probe schedules and late results

Probe deadlines make old work useless. Duplicate tasks, delayed results, and configuration edits must not change the wrong outage decision.

### Incident changes and alerts

An outage decision must survive crashes without duplicate incidents. An external webhook can succeed even when the sender receives no response.

### Public URLs and secrets

Customers control target and webhook URLs. DNS changes and redirects can expose internal services or send credentials to another origin.

## Common mistakes

- Treats a missing probe result as a target outage.
- Counts attempts instead of distinct probe regions.
- Combines different rounds or configuration versions in one decision.
- Replays stale probes after a location recovers.
- Claims exactly-once alerts without receiver cooperation.
- Protects probe destinations but ignores webhook destinations or credential exposure through redirects.

## The bar by level

- **Senior:** Defines timed rounds, configuration versions, unknown evidence, and durable schedules. Explains incident transactions, webhook retries, stale work, and URL controls for probes and webhooks. Prevents credential exposure through redirects.
- **Staff+:** Tests regional independence and health evidence under correlated faults. Defines behavior during control failures and delayed health reports. Quantifies detection limits, storage cost, and recovery load. Defines trust boundaries for destinations, credentials, and tenants.

Practice this question with an AI interviewer: https://sysround.com/practice/new?problem=synthetic-uptime-monitoring


---
SysRound (https://sysround.com) is an AI practice tool for system design interviews. It has a voice or typed AI interviewer and a shared whiteboard. At the end, you get feedback on each skill and a model answer. The AI feedback can be wrong.