Share this post
Cloud-managed edge runtimes are capable for containerized workloads. Azure IoT Edge, for example, is a natural starting point for organizations already invested in the Microsoft stack.
The question worth asking isn't whether it works today. It's whether the vendor plans to keep investing in it on the same timeline your operation depends on.
Those two timelines don't naturally align.
{{article-cta}}
Your operation plans in years. Cloud vendors don't.
Manufacturing and industrial operations don't cycle through infrastructure the way enterprise software gets updated. When a plant deploys containerized applications across production sites, they do so with the expectation that the management layer will still be there in year three, seven, and even year ten. Equipment decisions, certification processes, and planned upgrade windows are all built around that assumption.
Cloud IoT services don't carry that commitment. They exist while the economics justify building them, and they get cut when priorities shift. Google shut down its Cloud IoT Core service in 2023 after just five years of general availability, and it's not an isolated case. IoT Analytics tracked over 620 IoT platforms operating globally in 2019, and the exits since then, from enterprise vendors including SAP, Cisco, and IBM, reflect a consistent pattern. IoT services built on top of larger tech portfolios get cut when they stop generating returns.
A 12-month migration notice doesn't align with a 10-year production plan. When you've built deployments across dozens of sites, a forced migration is a significant project that wasn't in the budget.
Cloud-managed edge runtimes are also built for IT specialists. If the people managing devices day-to-day can't run the platform without outside help, the operational risk doesn't stop at vendor longevity. It extends to a skill set most industrial organizations don't have on staff.
The vendor's product roadmap and your production plan are not the same document. Build your management layer on your timeline, not theirs.
The costs that don't show up in the headline price
Cloud-managed edge runtimes like Azure IoT Edge typically use consumption-based pricing based on subscription tiers, message volume, and API calls. Each charge is small in isolation, yet across dozens of sites and hundreds of devices, accurate forecasting before you reach full production scale is difficult. By the time you know what it actually costs, you're already committed to the platform, and the switching cost is high.
Setup adds more. Initial deployment typically requires cloud specialists or external consultants. That's a project cost and often an ongoing one. Hardware is a variable, too, with cloud runtimes that generally need more capable devices to meet their requirements.
Portainer Edge runs on predictable per-node licensing. The agent needs approximately 10MB of RAM, so existing hardware is usually sufficient. And because OT teams can manage rollout and day-to-day operations through the interface without cloud infrastructure knowledge on hand, companies eliminate specialist contractors for deployment and external dependency for routine maintenance.
By the time you know what it actually costs, you're already committed to the platform. Know the number before you start.
Production doesn't wait for connectivity
Cloud-managed edge runtimes route their core operations through vendor infrastructure. Without an active connection to the cloud control plane, provisioning stops and updates queue, and you lose visibility into your fleet. In well-connected environments, that's not a practical concern.
At a site with intermittent connectivity, a planned OT/IT network separation, or limited connectivity windows, it is. A planned update that can't reach the cloud during a maintenance window means the window closes with the work undone. A new device that can't be provisioned remotely means a technician is waiting.
With Portainer Edge, the management layer runs inside your own network. New devices onboard without an external connection. One-Touch Onboarding provisions device fleets through a script with no manual configuration per device. Updates and configurations deploy locally. If connectivity drops, devices keep running their applications, the Edge Agent stays resilient behind NAT, and state syncs on reconnect. Air-gapped environments are fully supported without additional setup. If the Portainer control plane itself is temporarily unavailable, devices keep running the containers already deployed to them.
For teams managing Docker workloads across a mixed fleet, Portainer handles native Docker and Docker Compose lifecycle management directly on the device, without wrapping workloads in the module and manifest model used by platforms like Azure IoT Edge.
The architecture has to work when the cloud doesn't. That's not an edge case. That's the edge.
Making the call
Cloud-managed edge is a workable option if your operation is already invested in a single cloud ecosystem, your sites have reliable connectivity, and your team has the expertise to run the platform.
For plants with distributed or remote sites, OT-managed environments, or infrastructure decisions that need to hold for a decade, dedicated edge management removes the dependency on a vendor whose product timeline is unlikely to stay aligned with yours.
We've compared both options across different dimensions: strategic fit, cost of ownership, and technical realities.
Download the full comparison of Portainer Edge vs Azure IoT Edge.
Looking at Portainer Edge vs. Cloud-Based IoT Services like AWS IoT Greengrass? Check out this comparison.
