Rutba Vision: video surveillance for camera networks past the size one recorder copes with
First published on rutba.io . A handful of cameras is a product you buy. A hundred and fifty, writing terabytes a day, is an operations problem, and it fails the same ways each time. The disks fill and recording silently stops. The event database bloats until queries time out. A camera's clock drifts and its footage stops counting as evidence. And one box is the whole system, so one failure is an…
Rutba Vision is a video surveillance solution designed to handle large camera networks without the limitations of a single recorder. It addresses common issues such as disk filling, event database bloat, clock drift, and single-point failure.
The software runs the same node agent across all machines, with each site's nodes serving three roles: edge nodes hold cameras, relay connections, and manage live view and detection; analysis nodes run AI and codec work; and the core handles the Vision Bus, API, and console. A single site may run all three roles on one machine, while a large warehouse can distribute them across several.
Each site operates independently, recording and analyzing its own cameras and storing footage locally. The site connects to a hub at the headquarters over TLS, which stores events, incidents, tracks, thumbnails, and health data. In case of an outage, only the footage from the affected site is unavailable until the hub acknowledges the outbox.
Recording is done by copying each camera's stream into self-contained fMP4 segments, hashed with SHA-256 as they are written. Segments are written to a spool and renamed into place, ensuring that a crash never leaves half a segment unplayed. Detection, AI, indexing, and distribution can all be down, but footage is still recorded and catalogued.
Each edge node manages its own storage guard for retention and free space. Video is stamped in UTC as it arrives, so camera clocks are not required, and AI can focus on specific zones. Inference runs on CPU, GPU, or NVIDIA cards, with permissively licensed default detectors.
Search queries can be made in plain language, such as asking for a red car at the north gate after 10pm yesterday. Evidence that requires hold is never deleted, regardless of retention settings. Rutba Vision is not a VMS; it feeds feeds into existing video management systems. The hardware and footage remain the responsibility of the organization, and they can choose to run Rutba Vision themselves or outsource it to a hub.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
