Data Management & Monitoring at The National Library of the Netherlands
summary
The National Library moves from static logs to a live observability stack: Grafana + Mimir + Loki running on two bare‑metal Grafana servers with Postgres, three Mimir and three Loki instances, and Grafana Agent rolled out via an Ansible collection. It ingests 3.5 PB of data growing ~300–400 TB/year, replaces Splunk, and cuts costs by €80,000/year. Teams onboard with two variables in their infra code and get two ready dashboards for metrics and logs; alerting and tracing are planned but not yet implemented.
AI-written summary · about
chapters (7)
- 0:00National Library mission
- 5:06Why observability matters
- 8:17Grafana over Splunk choice
- 10:07On-prem Grafana landscape
- 13:18Developer dashboards
- 15:48Future alerting and tracing
- 17:34Results and next steps
related talks
53mGetting started with Grafana Loki, a modern logs database
44mIntroduction to Grafana Loki
28mFrom Fragmented Logs to Unified Insights: Scaling O11y
01h 02mThe Power of Metrics, Logs & Traces with Open Source
40mLarge-Scale Logging Made Easy
37mGrafana Loki: Like Prometheus, But for logs.
21mActian & Grafana Cloud: The Search for a Customizable Observability Tool
17mImplementing Open-source Observability within Maersk