Data storage (farm and edge stacks)¶

On a smart farm, data storage spans at least two familiar layers: short-lived streaming handoff from sensors and gateways, and durable structured storage for assets, history, and maps.

Messaging handoff (MQTT)¶

MQTT is a common pub/sub protocol between edge devices, brokers, and application services (for example after LoRaWAN or DIY LoRa bridges). Official specification and overview material:

• MQTT — specification (overview and versions)
• MQTT Version 5.0 — OASIS Standard

In this vault, field-to-broker patterns are collected under LoRa, MQTT, and gateway bridges.

Relational and spatial databases (PostgreSQL / PostGIS)¶

PostgreSQL is a general-purpose relational database; PostGIS extends it for geometry and geography—useful for parcels, boundaries, and asset locations.

• PostgreSQL documentation (current)
• PostGIS — documentation

Slide capture (complements the docs above): PostGIS Day 2019 — PostGIS 3.0 deep dive (PDF slides).

Background captures (not a substitute for the manuals above):

• PostgreSQL — system design interviews (Algomaster capture)
• PostgreSQL internals — query flow (Medium, part 1 capture)
• Geospatial PostgreSQL for enterprise — PostGIS deep dive (Perry Robinson capture)
• GiST spatial index in PostGIS (Medium capture)
• PostGIS — complete workflow (capture)

Farm management and lab-build context: Farm data, farmOS, and agriculture lab builds. Product entity: farmOS.

Backup / restore (self-hosted stacks) — compare granularities for PostgreSQL-class data and volumes: Backup strategy comparison — farmOS, homelab, PostgreSQL, containers, Restore and recovery tiers — homelab farm systems.

Related¶

• LLM Wiki pattern — how provenance and synthesis relate in this repository
• Domain content overview — “data” strand among land, networks, power, and time