SLA, SLO and MTTR: Measuring Network Operations Properly
“99.9% uptime” sounds reassuring, but it allows almost nine hours of downtime a year — and says nothing about how quickly problems are fixed or how often users suffer degraded service. Better metrics lead to better operations.
Know the terms
- SLA (Service Level Agreement): the contractual commitment, often with credits if missed.
- SLO (Service Level Objective): the internal target you aim for, usually stricter than the SLA.
- MTTD: mean time to detect — how long before you know there is a problem.
- MTTR: mean time to restore — from detection to service working again.
- MTBF: mean time between failures.
What to measure
- Availability per site and per service, not just averaged across the network.
- Incidents detected by monitoring versus reported by users — a high share from users means monitoring gaps.
- MTTR by incident type and by carrier.
- Repeat incidents on the same link or device.
- Performance: latency, packet loss and utilisation, not only up/down.
Report for decisions
A good monthly report highlights the worst sites, recurring causes and the actions taken — not just green dashboards.
Our Network Operations service reports all of these every month. See a sample report.
