Lead Forensics

How should airports plan resilient SAN storage for CCTV footage?

Airports should plan resilient SAN storage for CCTV footage by sizing capacity from camera count, bitrate and retention, separating recording traffic from analytics and evidence workloads, and removing every single point of failure across controllers, paths, power and drives. Failover behaviour must be tested, not assumed. This article covers capacity, architecture, failover, SAN versus hyperconverged options, analytics and evidence demands, and ongoing monitoring for airport SAN storage and surveillance servers.

How much SAN capacity does an airport CCTV estate need?

SAN capacity for airport CCTV equals the combined camera bitrate multiplied by the retention period, plus headroom for growth, RAID overhead and analytics data. The result is rarely close to a simple per-camera estimate, because bitrates vary with resolution, codec, frame rate and scene activity.

A practical calculation follows these steps:

  1. Group cameras by type, such as fixed terminal cameras, high-resolution apron or perimeter cameras, and ANPR cameras on forecourts and car parks.
  2. Assign a realistic average bitrate to each group, based on vendor data or measured samples from the actual scenes, not datasheet maximums.
  3. Multiply by recording hours per day (24 for continuous recording) and by retention days, including any longer retention required by policy or regulation.
  4. Add RAID or erasure protection overhead, hot spares and at least 20 to 30 percent growth headroom.

Treat any figure from a worked example as hypothetical until validated against your own cameras. Busy concourses and baggage halls produce higher bitrates than quiet back-of-house corridors, and a design based on averages can underestimate peak write demand. Capacity is also only half the picture. Sustained sequential write throughput across every simultaneous stream matters just as much, and a SAN that has enough terabytes but cannot sustain the aggregate ingest will drop frames. Secure Logiq’s server selection guidance explains how to relate stream counts and bitrates to server and storage requirements.

What SAN architecture keeps airport recording uninterrupted?

An uninterrupted airport SAN architecture removes single points of failure at every layer: dual controllers, redundant fabric paths, redundant power supplies on separate feeds, hot-spare drives and recording servers that can fail over to one another. Redundancy at only one layer leaves the system exposed at the next.

The main design elements are:

  • Dual controllers and multipathing: each recording server should reach storage through at least two independent paths, so a failed cable, switch port or controller does not interrupt writes.
  • Redundant fabric: two separate switches or fabrics, ideally in different rooms or risers, so a single local incident cannot isolate the array.
  • RAID with hot spares: parity-based RAID such as RAID 6 or RAID 60 tolerates drive failures while rebuilding onto a spare. Larger drives mean longer rebuilds, so protection levels should reflect drive size and array width.
  • Redundant power and cooling: dual power supplies fed from separate circuits, backed by UPS and generator supply, are standard expectations in airport equipment rooms.
  • Segregated recording traffic: keep recording streams off the network used for operator viewing and general IT traffic, so a busy control room cannot starve ingest.

For multi-terminal airports, splitting storage across physically separate equipment rooms limits the impact of a localised fire, flood or power fault. RAID is not a backup, and redundancy does not on its own guarantee uninterrupted recording, so the design must be assessed as a complete system. Secure Logiq’s system design service exists for this kind of end-to-end review.

How should airports handle SAN failure and failover?

Airports should assume components will fail and design so that a single failure never stops recording. This means automatic path and controller failover, VMS-level recording server failover, and a documented, rehearsed procedure for the failures that automation cannot cover.

What happens when a drive fails?

With RAID 5, 6 or 60 and a hot spare, the array continues to serve writes while rebuilding. Rebuild time and the performance dip during it should be known in advance, because a rebuild competes with live recording for the same disks. Alert thresholds should flag the failure immediately, and replacement drives should be held on site or covered by a fast-response warranty.

What happens when a controller, path or recording server fails?

Multipathing should shift traffic to the surviving path without operator action. For recording server loss, most VMS platforms offer failover recording servers, but behaviour differs by vendor, so confirm exactly how your VMS handles in-flight streams and where gaps in recording may appear. Where evidence continuity is critical, consider edge recording on cameras as a secondary safety net during the failover window.

Finally, test it. Scheduled failover exercises, such as pulling a path, powering down a controller or disabling a recording server, reveal configuration errors long before a real incident does. Disaster recovery planning, including what is replicated off site, should be a separate decision from local resilience.

Is a SAN or hyperconverged storage better for airport CCTV?

Neither is universally better. A dedicated SAN suits large airports with very high stream counts, long retention and a need to scale storage independently of compute. Hyperconverged or software-defined storage suits smaller or phased deployments that benefit from simpler scaling and consolidated hardware.

The trade-offs compare as follows:

  • SAN: storage scales independently, controllers and arrays are mature and well understood, and sizing for sustained sequential writes is straightforward. The downsides are a separate fabric to manage, more components and potentially a larger rack footprint.
  • Hyperconverged: compute and storage scale together in nodes, with fewer discrete components and often easier expansion. The downsides are that adding capacity may also add compute you do not need, and node-to-node replication traffic must be factored into the network design.

The decision should rest on stream count, retention, growth plans, in-house skills and how the VMS and analytics platforms are supported. Secure Logiq has described its Logiqal Core Architecture and Logiqal Video Volume as a unified compute and storage platform with a software-defined storage layer for sequential video workloads, as covered in its Logiqal Core Architecture announcement. Its Qatar University SAN storage case study shows a SAN-based deployment in practice. Compare each option against your specific requirements rather than defaulting to either.

How can storage cope with video analytics and evidence requests?

Storage copes with analytics and evidence workloads by isolating them from live recording. Analytics generates metadata and heavy read activity, while evidence export creates sudden bursts of reads. Without separation, either can reduce write performance and cause dropped frames.

Several design choices help:

  • Separate video and metadata: analytics metadata is small, random and frequent, unlike sequential video. Storing both on the same volumes increases fragmentation and drive wear, so keep them apart where the VMS and storage platform allow.
  • Use caching deliberately: SSD caching or tiering can absorb metadata and search activity, while high-capacity HDD arrays handle the bulk sequential video.
  • Size for reads as well as writes: forensic search, multi-camera playback and bulk exports during an incident can all coincide with peak recording. Allow headroom for them.
  • Protect evidence: define how footage is locked or copied to a protected volume, with access logging, so retention policies do not overwrite material under investigation.

Analytics workloads on GPUs and processing servers also depend on fast, consistent access to stored or live video, so analytics planning belongs in the original storage design, not as an afterthought.

What should airports monitor and maintain on CCTV storage?

Airports should monitor drive health, array and rebuild status, path and controller state, capacity and retention trends, write latency and recording gaps, and maintain firmware, spares and tested recovery procedures. Monitoring should alert operations staff in real time, not rely on periodic manual checks.

A practical maintenance checklist includes:

  • SMART and predictive drive failure alerts, with prompt replacement of degraded drives.
  • Alerts for failed paths, controllers, power supplies and fans, because a silently failed redundant component means the next fault causes an outage.
  • Capacity and retention reporting, so camera additions or bitrate increases do not quietly shorten retention below policy.
  • Verification that recorded footage is actually playable, including spot checks across cameras and time ranges.
  • Planned firmware and VMS updates in controlled windows, tested first on non-production equipment where possible.
  • Warranty and spares strategy, with response times matched to how critical the recording is.

Review the design whenever cameras, resolutions or analytics loads change. Storage that was correctly sized at commissioning can become a bottleneck after a terminal extension or a move to higher-resolution cameras.

Planning SAN storage for an airport CCTV estate is easier with an independent review of capacity, throughput and failure scenarios. To discuss your requirements, explore Secure Logiq’s system design expertise or contact the team directly.

Frequently Asked Questions

Where should I start if I'm planning a new SAN for an airport CCTV system?

Start with a camera inventory grouped by type, resolution, codec and frame rate, then capture measured bitrate samples from real scenes such as concourses, baggage halls and aprons. Confirm retention requirements from your security policy and any regulatory obligations, and agree growth assumptions for the next three to five years. With those inputs, you can calculate capacity and sustained write throughput and ask vendors to design against specific numbers rather than generic estimates.

How often should we test SAN failover in a live airport environment?

Run structured failover tests at least annually, and after any significant change such as a firmware upgrade, VMS update or controller replacement. Test one failure type at a time, such as pulling a path, powering down a controller or disabling a recording server, and check for gaps in recorded footage afterwards. Schedule tests in low-traffic windows, notify operations staff in advance, and record the results so you can compare behaviour over time.

What are the most common mistakes when sizing CCTV storage?

The most frequent errors are using datasheet maximum or average bitrates instead of measured ones, ignoring write throughput in favour of raw terabytes, and leaving out RAID overhead, hot spares and growth headroom. Another is forgetting that analytics metadata, evidence exports and forensic searches add read load on top of recording. Re-check your assumptions whenever cameras are added or upgraded to higher resolutions.

How can we tell if a failed redundant component has gone unnoticed?

Configure real-time alerts for failed power supplies, fans, paths, controllers and degraded RAID groups, and route them to a monitored operations console rather than relying on email alone. Many arrays and SAN switches support SNMP or syslog integration with central monitoring tools. Supplement this with periodic checks of multipath status on each recording server, since a silently failed path means the next fault could cause an outage.

What can we do to avoid losing footage during the failover window?

Confirm with your VMS vendor exactly how failover recording servers take over in-flight streams and how long the switchover takes. Where evidence continuity is critical, enable edge recording on cameras with SD cards or onboard storage so footage can be back-filled after failover. Test this by deliberately disabling a recording server and reviewing the timeline for gaps.

How do we make sure footage needed for an investigation isn't overwritten?

Define a clear evidence procedure that locks or copies relevant footage to a protected volume with restricted, logged access before the normal retention period expires. Most VMS platforms support protecting or bookmarking sequences, but check how this interacts with your storage capacity. Allocate dedicated capacity for protected evidence so locked footage doesn’t reduce the retention of live recordings.

Should disaster recovery be handled by the same SAN design as local resilience?

No, treat them as separate decisions. Local resilience such as dual controllers, multipathing and RAID keeps recording running through component failures, while disaster recovery addresses loss of a whole room or site. Decide what, if anything, needs to be replicated off site, such as evidence clips, key cameras or configuration backups, since replicating all footage is often impractical in cost and bandwidth. Remember that RAID is not a backup.

Related Articles