SAN (Storage Area Network) is high-performance storage presented as raw block devices over a dedicated storage network, typically Fibre Channel or iSCSI. Unlike a NAS, which serves files over the ordinary LAN, a SAN gives servers what looks like a local disk over a separate, fast fabric, with the throughput and low latency that heavy, continuous workloads need. It is the enterprise choice for large-scale live recording where many cameras write at once.
Its strength is sustained write performance and scale. Because it is block-level over a dedicated network, a SAN can absorb the relentless parallel write stream of hundreds or thousands of cameras without the protocol overhead that throttles file-level storage, and it centralises storage so multiple recording servers share a large, managed, resilient pool. This suits big deployments — city systems, large campuses, data-centre-grade VMS — where capacity and throughput both have to be high and manageable.
The pitfalls are cost and complexity. A SAN with its dedicated fabric, switches, and management is significantly more expensive and more specialised to run than NAS or direct-attached disks, so it is overkill for small or moderate systems where simpler storage would record the same cameras for far less. The sound approach matches storage class to load: direct-attached or NAS for smaller and archival needs, SAN reserved for high-density live recording that genuinely needs its throughput — and, as always, sized for sustained write plus RAID overhead, with redundancy, not just headline capacity.
