Glossary
The monitoring glossary.
Clear, jargon-free definitions of the uptime, reliability, and monitoring terms teams use every day.
Monitoring types
API monitoring
Testing API endpoints with real requests and asserting on status, timing, and response body.
Cron job monitoring
Watching scheduled jobs so a missed or failed run becomes an alert instead of a silent failure.
DNS monitoring
Watching DNS records and resolution so misconfigurations or expiries don't quietly take you offline.
Heartbeat monitoring
Alerting when a scheduled job or process fails to check in on time — ideal for backups and cron jobs.
Keyword monitoring
Checking that a page actually contains (or doesn't contain) specific text, not just that it loaded.
Multi-region monitoring
Running checks from several locations and confirming failures from more than one before alerting.
Ping monitoring (ICMP)
Checking that a host is reachable on the network and measuring latency and packet loss.
Port monitoring (TCP)
Confirming that a specific TCP port and service — like a database, mail, or SSH — is accepting connections.
Real user monitoring (RUM)
Measuring the experience of actual visitors as they use your site, using data collected in their browsers.
SSL certificate monitoring
Tracking TLS/SSL certificates and domains so you're warned before they expire or become invalid.
Synthetic monitoring
Proactively testing a service with simulated requests from outside, rather than waiting for real users to hit problems.
Uptime monitoring
Automatically checking that a service is online and alerting you the moment it goes down.
Website monitoring
Checking that a website loads, returns the right status and content, and responds quickly.
Reliability metrics
Availability
A measure of how reliably a service is reachable and usable, typically expressed as an uptime percentage.
Downtime
Any period when a service is unavailable or not functioning correctly for users.
Error budget
The amount of unreliability an SLO allows — the gap between your target and 100%.
MTBF (Mean Time Between Failures)
The average time a service runs between one failure and the next.
MTTD (Mean Time To Detect)
The average time between when a problem starts and when someone (or something) notices it.
MTTR (Mean Time To Recovery)
The average time it takes to restore a service after an outage begins.
Response time (latency)
How long a service takes to respond to a request — a key early-warning signal before outages.
SLA (Service Level Agreement)
A formal commitment to a level of service — often an uptime guarantee — with consequences if it is missed.
SLI (Service Level Indicator)
The actual measurement used to judge a service's reliability, such as the percentage of successful requests.
SLO (Service Level Objective)
An internal reliability target for a service, such as 99.95% uptime over 30 days.
The nines of availability
Shorthand for availability levels — 99.9% is 'three nines', 99.99% is 'four nines', and so on.
Uptime
The percentage of time a service is available and working, usually measured over a month or year.
Alerting & incidents
Alert fatigue
When too many alerts — especially false ones — cause responders to miss or ignore the important ones.
Escalation policy
Rules that decide who gets alerted, in what order, if an incident isn't acknowledged.
Incident management
The process of detecting, responding to, resolving, and learning from service disruptions.
On-call
The rotation of engineers responsible for responding to alerts and incidents at any given time.
Postmortem (post-incident review)
A structured, blameless review after an incident to capture what happened and prevent recurrence.
Status page
A public page that shows the current and historical status of your services to customers.
Concepts
Check interval
How often a monitor runs its check — e.g. every 30 seconds or every 5 minutes.
HTTP status codes
The numeric codes servers return to indicate the result of a request — 200 OK, 404, 500, and more.
Webhook
An automated HTTP callback that sends data to a URL when an event happens — used to wire alerts into other tools.
Put the theory into practice
Spectra turns these concepts into working monitors, alerts, incidents, and status pages. Free forever, no credit card.
