All case studies
ISP / Carrier · 2024
Standing up a 24/7 NOC and monitoring stack for a metro FNO
The operator could meet its SLAs technically but could not prove or defend them. We built the observability layer and the human process around it.
Metro fibre network operatorKwaZulu-Natal, South Africa
- -83%
- Alert noise volume
- 4 min
- Mean time to detect (from 26 min)
- 99.98%
- Verified core availability
- 24/7
- Staffed NOC coverage
The challenge
- Outages were discovered by customers before engineers.
- Alert noise from an unfiltered legacy tool meant real events were ignored.
- No SLA reporting for enterprise clients.
What we did
- Deployed Prometheus, SNMP exporters and syslog correlation with dependency-aware alerting.
- Built per-client SLA dashboards and automated monthly availability reporting.
- Defined severity levels, escalation matrices and shift handover runbooks.
- Ran tabletop incident drills with the new NOC team.
Technology
- Prometheus
- Grafana
- Graylog
- Netbox
- Alertmanager
Facing something similar on your network? We start with a measured assessment, not a quote for hardware.
Book an assessment