A centralised SAN consolidates video storage from many recording servers into shared, high-density arrays, which usually reduces total rack units and the number of separate power supplies and fans in a surveillance estate. The saving is real but conditional: the SAN, its switching and its controllers add their own space and power demands. This article covers the rack, energy, cooling and design questions that decide whether a SAN suits your surveillance servers.
How much rack space can a centralised SAN save?
A centralised SAN saves rack space by replacing many partly filled storage chassis with a few high-density arrays, but the saving depends on the design. No universal percentage applies, because the result varies with drive capacity, retention targets and how much unused capacity the current servers carry.
In a direct-attached design, each recording server carries its own drives sized for its worst-case camera load. Capacity that one server never uses cannot help another. A SAN pools that capacity, so storage is provisioned against the combined requirement of the whole estate rather than the sum of individual peaks. Fewer chassis, fewer drive bays sitting empty and fewer 2U or 4U servers in the rack follow from that.
The saving is not unlimited. A SAN deployment still needs:
- Storage controllers or array enclosures
- Switching or fabric components, such as Fibre Channel or iSCSI switches
- Recording servers, which still decode, write and run the VMS
A typical scenario might compare thirty recording servers, each with its own drive bays, against fewer recording servers attached to shared arrays. That is a hypothetical illustration, not a guaranteed outcome. A reliable figure comes only from a calculation based on your camera count, bitrate, retention period and RAID overhead.
Secure Logiq’s design team routinely examines whether a specification can be met with a smaller physical footprint, and its system design service includes the bandwidth and storage calculations needed to test that.
Does centralised SAN storage reduce energy consumption?
Centralised SAN storage can reduce energy consumption when it removes redundant servers, power supplies and idle drives, but it does not do so automatically. Savings come from consolidation and better capacity utilisation. They are offset by the power draw of the arrays, controllers and switching that a SAN introduces.
Each server in a direct-attached estate carries a fixed overhead: motherboard, memory, fans and redundant power supplies. Reducing the number of chassis removes that overhead many times over. Pooled capacity also means fewer spinning drives are needed to meet retention targets, because stranded capacity is no longer purchased and powered.
Drives remain the dominant consumers in a storage-heavy system. A SAN that needs the same number of drives as the servers it replaces will not save much drive power, only chassis overhead. Equally, recording servers still need enough CPU and memory to write and process video, so compute demand does not disappear.
Judge energy on a total-system basis. Compare the combined draw of recording servers, arrays, controllers and switches against the draw of the distributed alternative under realistic load. Check the efficiency ratings of the power supplies in each design, since redundant 80 Plus certified units are a common feature of enterprise surveillance hardware.
What cooling and power density issues does a SAN create?
A SAN concentrates heat and power into fewer racks, raising per-rack power density and cooling demand even when total consumption falls. Fully populated storage enclosures can draw significant power and need reliable airflow, so the facility must support the higher density before consolidation is worthwhile.
Many surveillance rooms and edge cabinets were sized for a spread of modest loads, not a few dense ones. Moving to a SAN can expose limits in:
- Power distribution per rack, including PDU capacity and circuit ratings
- Cooling capacity and airflow direction, particularly in hot and cold aisle layouts
- Floor loading, since drive-dense enclosures are heavy
- Start-up current, as many drives spinning up together create a surge
Noise is a further consideration in occupied buildings. Dense enclosures with high-speed fans suit a data room but not an open office or back-of-house hospitality space. Confirm the environment before specifying equipment.
Concentration also changes the failure picture. A cooling or power fault affecting one dense rack can impact a larger share of your recording than the same fault in a distributed design. Redundant power feeds, monitored environmental sensors and sensible rack placement help limit that risk.
What’s the difference between SAN and direct-attached storage for surveillance?
The key difference is that direct-attached storage (DAS) connects drives to a single recording server, while a SAN provides shared block storage over a dedicated network that multiple servers can use. DAS is simpler and self-contained. A SAN offers pooled capacity and flexibility at the cost of added complexity.
Direct-attached storage
Each server owns its drives, so performance is predictable and the failure domain is contained to that server. There is no storage network to design, monitor or secure. The drawback is stranded capacity: one server may sit near full while another has spare space it cannot share. Growth usually means adding another server.
Centralised SAN
Servers access shared volumes over a storage network, so capacity can be allocated and reallocated as needs change. It suits large, multi-server estates and environments where retention requirements vary between sites or camera groups. The trade-offs are the cost and skill needed to design the fabric, the dependence of many servers on shared components, and the need to ensure the SAN sustains continuous sequential write performance across all connected recorders.
Neither is inherently better. SAN is also distinct from NAS, which presents files over a network rather than block storage, and the right choice depends on the VMS and its storage requirements. For a real-world example of shared storage in a large deployment, see the Qatar University SAN storage case study.
When is a centralised SAN the right choice for surveillance?
A centralised SAN is the right choice when a surveillance estate is large enough that pooled capacity, simpler expansion and consolidated management outweigh the added complexity. It is usually less suitable for small or single-server systems, where direct-attached storage is simpler, cheaper and easier to maintain.
A SAN tends to make sense when:
- Many recording servers share a campus, city or multi-building estate
- Retention periods differ widely, leaving capacity stranded on individual servers
- Rack space, power or cooling in the equipment room is tightly constrained
- The organisation has IT skills to manage a storage network and, ideally, existing infrastructure
- Centralised, enterprise-grade resilience is a priority
Be cautious where the team lacks storage networking experience, where the site is small, or where a shared storage layer would become a single point of failure without proper redundancy. Remember that RAID is not a backup, and redundancy does not by itself guarantee uninterrupted recording.
Whichever route you choose, size the system around sustained throughput, retention and growth, not capacity alone. The server selection guide and Secure Logiq’s range of HD surveillance servers are useful starting points for comparing architectures.
If you are weighing SAN against direct-attached storage for a specific project, speak to Secure Logiq’s system design team about modelling rack space, power and retention for your particular camera estate.
Frequently Asked Questions
How do I calculate the storage I need before deciding between a SAN and direct-attached storage?
Start with your camera count, the average bitrate per camera (in Mbps), the required retention period in days and the recording mode (continuous or motion-based). Multiply these to get raw storage, then add RAID overhead and a growth margin of 20-30%. Compare the result for a pooled SAN design against the sum of individual server requirements to see how much stranded capacity you could recover.
What should I check about my equipment room before moving to a SAN?
Survey rack-level power (PDU capacity and circuit ratings), cooling capacity and airflow direction, floor loading and available space for storage enclosures and switches. Also check the in-rush current when many drives spin up together and whether you have dual power feeds. If any of these are marginal, you may need facility upgrades before consolidating.
How can I avoid the SAN becoming a single point of failure?
Use redundant controllers, dual power supplies on separate feeds, and multipath connectivity through at least two independent switches or fabrics. Monitor drive health, temperature and fabric status with alerts, and keep hot spares in the arrays. Remember that RAID is not a backup, so consider whether critical footage also needs a separate copy or failover recording path.
Should I choose Fibre Channel or iSCSI for a surveillance SAN?
Fibre Channel offers dedicated, low-latency transport but needs specialist switches, HBAs and skills. iSCSI runs over Ethernet, which is often cheaper and easier to manage if your team already knows IP networking, but it needs a properly isolated, adequately provisioned network. Check your VMS vendor’s supported storage protocols and validate sustained write throughput before committing.
What are the most common mistakes when deploying a SAN for video surveillance?
The most frequent errors are sizing for capacity alone and ignoring sustained sequential write throughput, under-specifying the storage network, and sharing the SAN with unrelated enterprise workloads that compete for performance. Others include skipping redundancy in power and switching, and not planning for growth. Testing with a realistic load from all recorders before go-live helps catch these issues.
Can I migrate from direct-attached storage to a SAN without losing existing recordings?
Usually yes, but it needs planning. Common approaches include running the new SAN in parallel and letting the old DAS age out under your retention policy, or copying archives to the SAN where the VMS supports it. Consult your VMS vendor’s migration guidance, schedule changes outside peak recording periods, and keep a rollback path until the new storage has been proven under full load.
Is a SAN worth it for a smaller surveillance system?
Generally not. For a single server or a handful of recorders, direct-attached storage is simpler, cheaper and easier to maintain, and the pooling benefits of a SAN are limited. Consider a SAN once you have many servers, widely varying retention needs or serious constraints on rack space, power or cooling, and you have the skills to manage a storage network.
Related Articles
- When does a CCTV deployment need petabyte-scale SAN storage?
- How should airports plan resilient SAN storage for CCTV footage?
- What factors affect surveillance server storage requirements in 2026?
- What is the difference between M series and H series surveillance servers?
- Are surveillance servers scalable for future camera expansion?