VMS architecture questions come up in nearly every enterprise security evaluation, and for IT teams, the conversation often moves fast. Bandwidth requirements, data residency, integration dependencies, what the system looks like when the internet goes down: these are the details that shape how a deployment actually performs, and they’re worth working through before direction gets set.
Working through these evaluations regularly, the pattern is consistent: organizations that approach cloud versus hybrid VMS with a structured set of infrastructure questions make better decisions and avoid the surprises that show up after deployment. That framework starts with infrastructure considerations, the right questions for any VMS evaluation, and a clearer picture of what hybrid architecture means beneath the marketing language.
Why Bandwidth Is the First Number IT Should Calculate
VMS bandwidth requirements are one of the most consistently underestimated factors in cloud deployments. It’s one of the most common gaps in VMS evaluations and one of the most consequential. Every other consideration of data residency, failover, or integrations assumes the network can carry the load. Bandwidth is the foundational question that the others sit on top of.
A mid-size deployment of 200 cameras running at 3 Mbps each generates roughly 600 Mbps of continuous upstream traffic. The architecture decision shapes what happens to that load:
Cloud VMS — all video streams to the vendor’s infrastructure continuously; full WAN capacity required at all times.
Hybrid VMS — video records and stays on local infrastructure; the cloud serves as a secure connection point for remote access and enables administrators to view system health, licensing data, and other management information without that data ever leaving local infrastructure.
Most enterprise networks aren’t provisioned for sustained 600 Mbps outbound. The hybrid model lets IT evaluate against existing capacity rather than planning a parallel infrastructure upgrade.
Before committing to an architecture, model bandwidth across three scenarios:
- Normal operations — sustained upstream load across all active cameras
- Security event — concurrent streams, clip exports, and remote access traffic
- Network degradation — what remains functional locally if WAN capacity drops
The goal isn’t to predict every scenario perfectly; it’s to surface gaps early enough to address them before architecture is locked in. That analysis is often the difference between a deployment that fits the environment and one that requires the environment to change.
When Cloud-Hybrid Makes Sense and When It Doesn’t
Cloud VMS makes sense in specific circumstances: when outbound bandwidth is effectively unlimited, when the deployment is small enough that continuous upstream traffic isn’t a meaningful constraint, or when upfront hardware costs aren’t feasible and a subscription model is the only viable path.
Outside those conditions, the tradeoffs start to work against cloud VMS. Bandwidth costs climb as deployments scale. Data leaves the local environment continuously. And the subscription model that looks cost-effective at the start of a contract can look different as the deployment grows and pricing scales with it.
Hybrid VMS handles most of those tradeoffs differently. Data stays local by default. Remote access works without special network configuration. An open platform means cameras, servers, software, and analytics integrations aren’t tied to a single manufacturer; each piece can be selected, swapped, or upgraded independently. Changing VMS platforms doesn’t mean replacing cameras.
CompleteView is built on that model. Every deployment starts on-premises: local servers keep high-bandwidth video data within the local environment, and the customer retains full control over where data is stored and who can access it. Cloud Services is an optional add-on that creates the hybrid connection – secure remote access to system health metrics and live and recorded video without special network configuration. When a remote user connects, they access the system directly through the cloud connection point, video stays on local infrastructure and is never stored or transferred to the cloud
What Happens When the Internet Goes Down
For security operations, an internet outage isn’t an inconvenience; it’s a test of whether the system was built to operate independently or built to depend on connectivity.
In a fully cloud-dependent architecture, that test has a predictable outcome. Recording stops. Live monitoring goes dark. Any footage captured before the outage sits in vendor-managed storage that may be inaccessible until connectivity is restored. For an organization trying to respond to an active incident or reconstruct one afterward, that gap has real operational consequences.
Hybrid VMS handles the same scenario differently. Because recording happens on local infrastructure, an internet outage doesn’t interrupt local operations. Cameras keep recording. Security staff with access to local clients keep monitoring, and footage remains accessible on-site without waiting for connectivity to be restored. Remote access via web or mobile clients will be unavailable during the outage, but the system keeps running, and nothing is lost.
Cloud connectivity in a hybrid model adds genuine value — remote access, centralized management, multi-site visibility, but the system doesn’t depend on it. When the internet comes back, those capabilities resume. In the meantime, nothing stops.
The question worth pressure-testing with any VMS vendor: walk me through exactly what happens on-site, step by step, during a complete internet outage. The specificity of that answer tends to reveal how the architecture actually handles dependency.
The Questions IT Should Ask Before It’s Too Late
VMS decisions tend to have long tails. The system selected today will likely be in place for seven to ten years, and the constraints that aren’t visible in year one have a way of surfacing in year three. These are the questions worth working through before architecture gets locked in — they don’t require deep VMS expertise, just the same due diligence IT applies to any enterprise software decision.
- Camera ownership and portability: Who actually owns the camera hardware? Some cloud VMS platforms retain ownership of devices, meaning they’re non-transferable if the contract ends and in some cases vendors can remotely disable devices entirely. And if you discontinue the service, can those cameras be used with another platform? Confirm ownership and portability in writing before procurement.
- Data ownership: Where is video stored, who controls it, and how is it accessed if the service is terminated? Archived footage locked in a proprietary format or vendor-controlled storage creates significant exposure when switching providers.
- Video quality tradeoffs: Is the system making any compromises on video quality to keep cloud traffic feasible? Some platforms default to motion-only recording or reduced-resolution archives to manage bandwidth. Understanding what the system actually records — versus what it’s capable of recording — is a question worth asking explicitly.
- Service uptime guarantees: What are the contractual uptime commitments, and what recourse exists when they aren’t met? For organizations with genuine security requirements, availability guarantees need to be specific and enforceable.
- Data security: How is video secured at rest and in transit on the vendor’s infrastructure? Who has access to it, and under what conditions?
- Integration flexibility: What APIs are available for access control, alarm systems, and business intelligence platforms? Understanding whether integrations require standard API calls or vendor professional services changes the total cost picture significantly.
- Licensing at scale: How does pricing change as the deployment grows? Per-camera, per-site, or per-user models can create meaningful cost exposure that isn’t visible in an initial quote.
The pattern across these questions is the same: the risks that are hardest to see during an evaluation are the ones that are most expensive to discover after deployment.
The Value IT Brings to This Decision
The organizations that end up with VMS deployments that work well for IT and for the rest of the organization tend to be the ones where IT was part of the evaluation, not just the implementation.
That doesn’t require IT to lead the decision. It means coming to the conversation with the bandwidth model, the availability requirements, the integration dependencies, and the questions about data ownership already prepared. That preparation tends to lead to better decisions for the whole organization.
The teams that avoid the costliest VMS mistakes aren’t necessarily the ones with the deepest security expertise. They’re the ones who understood the infrastructure implications before the contract was signed; not the ones who inherited an architecture.
