Moodle Observability
Real monitoring of Moodle: metrics, logs, traces, and actionable alerts.
The problem
Without visibility, Moodle “goes down by surprise”. Observability allows anticipating degradation before it affects users.
What we implement
1. Infrastructure Metrics
- CPU, RAM, Disk, IO, Network per node.
- PHP-FPM: workers, queues, latencies.
- Redis: memory, hits/misses, latency.
2. Database
- Slow query log and critical query analysis.
- Connections and locks: bottleneck detection.
- Size and growth: capacity planning.
3. Moodle and User Experience
- Response times for key routes (login, course, quiz).
- Application errors: logs correlated by timestamp.
- Alerts: defined thresholds and notification per channel.
Result
- Fewer “mysterious” crashes.
- Data-driven decisions to optimize performance.
- Reduced MTTD/MTTR (faster detection and recovery).