Orateur
Description
RAPIDS: A Ready-to-Use EPICS Stack for Facilities Without a Controls Team
Author: Mauro Giacchini · INFN Laboratori Nazionali di Legnaro
Starting a new experiment or beamline with EPICS is harder than it looks. Installing EPICS base is straightforward — but a facility that is truly operational also needs data archiving, an electronic logbook, a pre-configured operator client, NTP synchronisation, and the safety tooling to keep everything running reliably under real conditions. For large accelerator facilities this stack is assembled over years by dedicated controls engineering teams. For a small experiment, a new beamline, or an R&D group adopting EPICS for the first time, that expertise and that time simply do not exist.
RAPIDS — Rapid Automated Production Infrastructure Deployment System — is designed precisely for this situation. Built on the community EPICS Training VM as its Ansible foundation, RAPIDS delivers a complete, production-ready operational stack in a single command:
- EPICS base and key modules (pvxs, asyn, StreamDevice, OPC-UA, autosave)
- Archiver Appliance 2.0.7 with configurable retention policies
- Olog 5.1.2 electronic logbook (ElasticSearch 8.11 + MongoDB 7.0)
- Phoebus pre-configured for both data browsing and logbook access
The entire deployment is driven from a single local.yml file that is the unique source of truth. A new facility can go from a vanilla AlmaLinux 9 installation to a fully operational EPICS environment — with archiving and logbook running — in under an hour, with no controls engineering background required.
RAPIDS is already in production at INFN Laboratori Nazionali di Legnaro, deployed for the ANTHEM collaboration between INFN-LNL and the University of Caserta, and currently undergoing validation for adoption on the SPES exotic-beam facility.
The architecture is explicitly designed for reuse: every component is parameterised, nothing is hardcoded, and enabling or disabling the archiver, the logbook, or specific EPICS modules is a one-line change.