What is observability?

In practice

The practical test is whether you find out about a failure before your customer does. That needs alerts on the things that matter to the business — a queue that stopped moving, a nightly job that did not run, an integration that returned errors all morning — rather than only on servers being up.

Where it fits in a build

Any system people depend on. It is the cheapest insurance in software and the first thing cut under deadline pressure, which is why so many outages are discovered by email.

Common mistakes

• Alerting on everything, so alerts are ignored.
• Logging without correlation identifiers, so a single case cannot be followed across systems.
• Monitoring uptime but not correctness — the system is running and has been producing wrong results since Tuesday.

Related terms

• Audit trail
• System integration
• Idempotency

Working on something that involves this?

ASTACKRA designs and builds AI systems, automation and custom software for businesses that need technology shaped around their own workflow. If this term turned up in a proposal and you want a straight answer about whether it applies to your situation, ask us — no obligation, and we will tell you if the answer is no.