CCTV footage can be replicated between separate sites by recording it at the source and copying it to a second location, either by streaming from the cameras to both sites at once or by copying recorded video between VMS recording servers over a WAN or VPN link. The right method depends on bandwidth, retention needs and how quickly the second site must hold usable footage.
Replication protects against theft, fire and site failure, but it must be designed around link capacity, security and failure behaviour. The sections below cover the main methods, bandwidth, security, hardware and resilience for an IP CCTV camera system.
Which replication methods work for multi-site CCTV?
The main replication methods are dual streaming, VMS-based archive replication, failover recording and storage-level replication. Each copies video to a second site in a different way, and most professional IP CCTV camera system designs combine two of them rather than relying on a single approach.
- Dual streaming: the camera or local recorder sends one stream to the local server and a second to the remote site. Both copies are live, but the remote link carries the full stream load continuously.
- VMS archive replication: the CCTV software copies recorded footage from one recording server to another, often on a schedule. This lets you move footage at quieter times, so the link does not need to carry peak load.
- Failover recording: a standby recording server takes over when the primary fails. This protects continuity of recording but does not always keep a full duplicate archive, so check what your VMS actually retains.
- Storage-level replication: the storage layer copies data between arrays, independent of the VMS. It suits large estates but needs careful design for sequential video workloads.
Feature names and behaviour vary between VMS platforms, so confirm the exact capabilities of your chosen CCTV software before committing to a design. Replication is also not the same as a backup: if corrupted or deleted footage is replicated, the damage is copied too.
How much bandwidth does replicating HD video need?
Bandwidth needed equals the combined bitrate of every stream you replicate, plus headroom. A single HD camera commonly runs at a few megabits per second, so ten replicated cameras may need tens of megabits per second, sustained, in both live and scheduled copy scenarios.
To estimate, multiply the average bitrate per camera by the number of cameras being replicated, then add around 20 to 30 per cent for protocol overhead and bursts. As a hypothetical example, 20 cameras at 4 Mbps each give 80 Mbps, or roughly 100 Mbps with headroom. This is an illustration of the method, not a guaranteed figure, because real bitrates depend on resolution, frame rate, codec, scene complexity and motion.
Several design choices reduce the load:
- Replicate only critical cameras rather than every stream.
- Replicate motion-triggered footage or event clips instead of continuous recording.
- Use efficient compression such as H.265 where cameras and VMS support it.
- Schedule archive copies for off-peak periods, provided the retention window allows it.
Remember that upload speed at the source site is usually the limiting factor, not download speed. Many business connections are asymmetric, so check the uplink figure carefully. Sizing tools such as the Logiqal Benchmark utility described in Secure Logiq’s resources help designers model stream loads before deployment, and the server selection guidance covers related sizing principles.
How is replicated CCTV footage kept secure?
Replicated footage is kept secure by encrypting it in transit, restricting access to it at rest, and isolating the replication path from general network traffic. At minimum this means a VPN or encrypted tunnel between sites, strong authentication, and role-based access control on the receiving server.
Practical measures include:
- Encrypted transport: use a site-to-site VPN or TLS-secured VMS communication so footage cannot be read on the wire.
- Network segmentation: keep surveillance traffic on a dedicated VLAN or separate network from corporate systems.
- Access control: apply least-privilege accounts, multi-factor authentication where supported, and audit logging on both sites.
- Hardened servers: keep operating systems and VMS software patched and disable unused services on the receiving server.
Data protection obligations also matter. Footage containing identifiable people is personal data, so UK organisations should consider UK GDPR requirements, including where footage is stored and who can access it. If the second site is in another country, review transfer rules before replicating. Secure Logiq hardware is manufactured in the UK, which can simplify supply chain and compliance conversations for security-sensitive deployments.
What hardware is needed at each site?
Each site needs a recording server sized for its own cameras, and the receiving site needs additional storage and processing capacity for replicated streams. The receiving server must handle its local recording load and the incoming replicated load at the same time, which is the most common sizing mistake.
At the source site, you typically need:
- A recording server with enough throughput for local cameras plus the outbound replication task.
- Resilient storage, such as hardware RAID with hot spares, so local recording survives a drive failure.
- A router or firewall capable of sustained encrypted tunnel throughput.
At the receiving site, you need:
- A server or storage array sized for the replicated retention period, in addition to any local recording.
- Dual power supplies and redundant storage where footage is critical.
- Sufficient network capacity to ingest replicated streams without contention.
Purpose-built surveillance hardware suits this workload because replication adds continuous sequential write and read activity that generic IT servers are not designed for. Secure Logiq’s HD surveillance servers are built for continuous video recording, and its bespoke solutions service covers multi-site requirements that standard configurations do not fit.
What happens if a site or link goes down?
If the link fails, local recording at each site should continue uninterrupted while replication pauses and resumes later. If a whole site fails, the replicated copy at the second site becomes the surviving record. The outcome depends entirely on how failover and catch-up are configured.
When the WAN link fails
Cameras should always record to local storage first, so a lost link causes a replication gap, not a recording gap. Check whether your CCTV software queues unsent footage and back-fills it once the link returns, or whether it simply skips the missed period. Back-fill traffic also competes with live replication, so the link needs spare capacity to catch up.
When a site fails completely
If a site is lost to fire, theft or power failure, the remote copy is your evidence. Test that the remote site can actually play back and export replicated footage, not just store it. Where live monitoring must continue, a standby or failover recording server and a clear process for redirecting operators are essential. Redundancy does not automatically guarantee uninterrupted recording, so document and rehearse the failure scenarios.
Whichever method you choose, treat replication as one layer in a wider resilience plan alongside RAID, power redundancy and a genuine backup strategy.
If you are planning replication across multiple locations, Secure Logiq’s system design service can help you model bandwidth, storage and failover requirements before you commit to hardware.
Frequently Asked Questions
How do I get started with replicating CCTV footage between two sites?
Start by listing which cameras are critical, the retention period you need at the second site, and the upload speed available at the source site. Then confirm which replication features your VMS supports (archive replication, failover recording or dual streaming) and pick one primary and one secondary method. Pilot with a small number of cameras over a VPN before rolling out to the whole estate, and measure real bitrates rather than relying on datasheet figures.
Can I replicate footage over a standard business broadband connection?
Often yes, but only for a limited number of cameras. Many business lines are asymmetric, so the upload speed is the real limit, and it must carry the combined bitrate of all replicated streams plus 20 to 30 per cent headroom. If the uplink is too slow, replicate only critical cameras, send motion-triggered clips instead of continuous video, or schedule archive copies for off-peak hours.
How can I test that my replicated footage is actually usable?
Run a regular test where you search for, play back and export a specific time period from the remote site’s copy, without using the primary server. Check that timestamps match, video is not corrupted and that exported clips open in a standard player or your evidence workflow. Schedule this check monthly or quarterly and record the results, so you find problems before you need the footage.
Does replication replace the need for a backup?
No. Replication copies changes as they happen, so deleted, overwritten or corrupted footage can be copied to the second site as well. A true backup keeps independent, point-in-time copies that you can restore from. Use replication for site resilience and combine it with RAID, power redundancy and a separate backup or archive policy for important evidence.
What are the most common mistakes when setting up multi-site CCTV replication?
The most frequent errors are under-sizing the receiving server so it cannot handle local recording and incoming replicated streams together, ignoring upload speed limits, and not checking whether the VMS back-fills footage after a link outage. Others include leaving replication traffic unencrypted, failing to monitor replication status and never testing a failover. Build in headroom, enable alerts for replication failures and rehearse your failure scenarios.
What should I do if replication falls behind or the link keeps dropping?
First check whether the link is saturated by comparing sustained bandwidth use against the available uplink, since back-fill traffic competes with live replication. If it is, reduce the replicated stream count, lower bitrates with H.265 or adjusted frame rates, or move to scheduled off-peak copying. If the link is unstable rather than overloaded, review the VPN configuration and consider a more reliable or secondary connection, while confirming that local recording continues unaffected.
Do I need to consider data protection rules when storing footage at a second site?
Yes. Footage of identifiable people is personal data, so UK organisations should consider UK GDPR requirements for storage location, retention periods, access controls and audit logging at both sites. If the second site is in another country, review the rules on international data transfers before replicating. Document who can access the replicated footage and apply the same retention and deletion policy at both locations.
Related Articles
- Why separate video recording servers from centralised storage?
- Which storage health indicators should CCTV operators monitor?
- How should airports plan resilient SAN storage for CCTV footage?
- What RAID level is best for HD surveillance servers?
- Should you choose a surveillance server or an NVR for enterprise use?