Skip to content

Storage backends

The writer speaks HTTP and supports three backend modes.

Fresh database required

Version 2.0.0 starts with fresh database schemas (QuestDB downsampling schema v1) and identifies home-automation devices by their AIN in the id tag; name is only a display tag. Point it at a new QuestDB instance, InfluxDB database or bucket. Existing data is neither migrated nor deleted.

Choose one backend through DB_TYPE. The backend hostname is mandatory; the mode-specific configuration is validated at startup. A full https:// hostname derives TLS and a port when no explicit port is set. Port 443 also implies TLS.

Recommended backend: QuestDB

Only QuestDB offers server-side downsampling with materialized rollup views, managed TTL retention, the fritzflux-cli admin tool, and Grafana dashboards that switch between raw data and rollups. The bundled Compose file also starts QuestDB. DB_TYPE defaults to influxdb_v1, so select QuestDB explicitly.

Backend differences

InfluxDB v1 InfluxDB v2 QuestDB
Protocol / port InfluxDB HTTP API, 8086 InfluxDB HTTP API, 8086 Influx Line Protocol over HTTP, 9000
Credentials Optional username/password, database Organization, bucket, token Optional HTTP basic authentication
Use when Existing legacy installation Current bucket/organization model New installations (recommended)
Dashboards influx1_dashboards/ influx2_dashboards/ questdb_dashboards/
Write precision ns ns s
Metrics Measurement fritzbox_<serial> Same as v1 Table fritzbox_<serial>_metrics, see QuestDB
Log and call entries Metrics measurement; sub-second offset from the entry uid Same as v1 Table fritzbox_<serial>_events with DEDUP, see QuestDB
Retention Configure in InfluxDB Configure in InfluxDB QUESTDB_DATA_RETENTION_DAYS, optional downsampling
fritzflux-cli No No Yes

The project does not create InfluxDB retention policies. Configure retention in the database platform according to your storage requirements.