Answers to the questions we hear most

For containerized workloads, modify your CI/CD pipeline to optionally push the container images to your free private Control Plane image registry. You instruct Control Plane to run your workload, either by specifying YAML, JSON, Terraform, or Pulumi script, or even by calling the RESTful API. Your workloads run inside of your Global Virtual Cloud (GVC™) – a named set of locations.

For non-containerized applications, Control Plane offers auto-containerization via the CLI plus a white-glove migration services if desired, to ensure optimal performance from day one.

Any workload can be containerized. If the automated tooling doesn't automatically create the deployment manifest for your workload, we provide free containerization services at no cost or obligation. Simply email support@controlplane.com and an experienced engineer will be glad to answer all your questions.

The only requirement is a Linux-compatible workload. Any Linux-compatible workload can run on Control Plane. The image can target Intel/AMD64 or ARM architecture. We also support virtual machines – so you can run ANY sort of workload, either Windows or Linux.

Yes. We offer a generous free trial period. Email sales@controlplane.com and we’ll set up a trial period that works for you.

Yes, absolutely. You can run workloads on any of your own cloud accounts, in addition to clusters provided in Control Plane cloud accounts. You can run workloads solely within your cloud accounts, solely in Control Plane cloud accounts, or a combination thereof. In addition, you can also run workloads on premises, using any bare metal or virtualized environment. Control Plane MK8s (Managed Kubernetes Service) can “Kubernetify” any VM or bare metal server, on any cloud or any bare metal brand. Whether you utilize an existing K8s cluster, or use MK8s to Kubernetify the servers, you Control Plane-ify cluster(s) with a single command and the whole Control Plane power apply to the one, or multiple K8s clusters.

Control Plane seamlessly integrates with any observability tool, although best-in-class observability is provided. In addition, any CI/CD tool, security scanning tools, any external container registries, and virtually all DevOps tools one can think of like Helm, Terraform, Pulumi, Cross Plane, CDK8s and pretty much anything else. Control Plane is API-first and thus, easily integrable with virtually anything that is programmable.

Yes. Many customers have containerized (with or without our help) monolithic applications that were previously deployed within virtual machines, or bare metal servers. In addition, we’ve helped several customers in breaking down their monolithic apps into smaller components that can be scaled and managed independently.

Yes. Running Windows applications requires using the VM workload type. We do not support containerized Windows workloads, but it is very straight forward to spin up Windows VMs and lift and shift existing Windows boxes – anywhere.

It depends on the size of the image, but there’s no delay beyond the physical speed in which images are pulled and executed, depending on the CPU. The performance is snappy and is dependent on image size, and app-specific initialization flow.

A workload is comprised of one or multiple containers. Since under the cover, Control Plane manages a fleet of Kubernetes clusters, a workload represents a pod. Inner workload/pod communication is facilitated on localhost. Cross workload communication is extremely flexible and uses mutual-TLS as the default. Service discovery is automatic and granular controls support any use case for complex workload to workload communication, within a single cluster/region, or across multiple clusters/regions/providers.

For blue/green releases, teams run the current version and the new version as separate Workloads in the same GVC. The new version can be deployed, scaled, and validated before it receives production traffic. Cutover is done by updating the Domain route to point from the blue Workload to the green Workload, which shifts traffic across the GVC without downtime. Rollback uses the same mechanism in reverse.

For canary releases, a Domain route can split traffic between the primary Workload and one or more canary Workloads using explicit weights. Teams can start with a small percentage of traffic on the new version, monitor behavior, then increase the weight until the new version receives 100% of traffic. Control Plane also supports mirrored traffic for pre-production validation before promotion.

All of Control Plane’s capabilities are provided through its API. Organizations can easily wrap the Control Plane API with interfaces that suit their requirements. Our professional services organization can work with clients to help customize capabilities as required.

Control Plane delivers all the benefits of AWS’ Lambda, but without the drawbacks. These drawbacks include:

• Only running in a single region.
• Amount of CPU/RAM must be pre-determined before deployment. It is not dynamic.
• Scale to zero introduces a cold start slowness.
• Handling cached data is non-deterministic and often does not work.
• Connection pooling is nearly impossible.
• Running long running transaction is not possible.
• Utilizing GPUs is not practical.
• Application logic is too fragmented, increasing complexity and maintenance headaches.
• More expensive than it ought to be.
• Observability is not included and is cumbersome to set up and expensive to maintain.
• Only a subset of programming languages and framework/versions are available.

Unlike the proprietary APIs of EKS, GKE, and AKS, Control Plane’s MK8s (Managed K8s Service) provides a singular API to deploy K8s clusters anywhere (e.g., AWS, GCP, Azure, Oracle, Linode, Hetzner, on-prem, VMWare, RaspberryPi, etc.) In addition, MK8s provides many add-ons that turn compute anywhere into enterprise-ready secure execution environments.

There are many significant differences. Let’s list the main ones:

• No need to provision whole servers. Applications dynamically scale horizontally and vertically while charges are calculated by millicores of compute.
• Private networking does not cost $2,000 per month. In fact, it is completely free.
• Workloads are not confined to running in a single region of AWS. Instead, workloads can work across any number of regions of any cloud. This is in order to provide high availability and low latency.
• Observability is not an expensive add-on, but rather, a basic feature of the platform.
• Workload to workload mutual TLS is built into Control Plane. A similar concept does not exist on Heroku.
• Control Plane provides Universal Cloud Identity, a capability allowing a workload running anywhere to mix-n-match ANY service of AWS, GCP, Azure and OCI. This is a unique capability of Control Plane that is not provided by Heroku, or any other hosting provider. A workload on Control Plane can simultaneously consume Big Query on GCP, S3 and Dynamo DB on AWS, and Cosmos DB and Active Directory on Azure, regardless if it is running on-prem, or any other cloud provider.

Control Plane provides 24/7 expert support provided by experienced DevOps engineers who are also contributors to open-source projects such as Kubernetes and Istio. Customer get Slack channels and typical response is < 5 minutes with a guaranteed response within a 30 minute window.

Control Plane’s SLA is 99.999%. Your workloads’ compute guarantee is 99.999% assuming you adopt at least two replicas in at least two locations. When you configure the Global Virtual Cloud (GVC™) simply choose more than one location. If a location becomes inoperable, the Control Plane system automatically routes requests to surviving healthy clusters.

Control Plane replicates its metadata across multiple geographic locations and automatically back up the data to a different region. Each quarter, Control Plane personnel perform a “table top” exercise, simulating data loss and exercising the restore procedures. Database and object and block storage data backup are the responsibilities of customers. Control Plane often provides best practices and advice, but clearly states that its responsibility is metadata (data about workloads and workload management). Any data backup and restore procedures are the responsibility of customers.

Yes. It is 99.999% availability and 30 minutes or less until support is acknowledged by real support engineers who actively work on a support case.

Everything is subject to audit trail. Whether leveraging the UI, CLI, API or Terraform/Pulumi other IaC products and the MCP Server, since everything ultimately enters the system through an API call, everything is subject to audit trail. Audit trail is kept forever. There is no purging of audit trail data and there is no charge for it.

Customer data is stored in either cloud/hosted databases or object/block storage. In this case, Control Plane does not store and manage application data. Control Plane does store metadata which is used in managing workloads and connection to data sources. In addition, Control Plane provides stateful workload support, in which workloads can attach persistent volumes. Control Plane provides backup facilities that are controllable via API, CLI, UI in order to facilitate data backup and restore processes.

Control Plane approaches security through defense in depth: a combination of technical controls, operational processes, third-party validation, and ongoing internal review.

Control Plane maintains SOC 2 Type II compliance and supports customers with security and compliance requirements across areas such as PCI and GDPR. Data is encrypted in transit and at rest, and access to systems is governed by authentication, authorization, and least-privilege principles.

Control Plane also conducts comprehensive third-party penetration testing twice per year. In addition, security reviews, monitoring, and internal controls are used to identify risks, detect suspicious activity, and improve the platform’s security posture over time.

Security is treated as an operational discipline, not a one-time certification exercise. Control Plane combines external validation, encryption, access control, monitoring, employee security awareness, and regular review processes to help customers protect sensitive workloads and meet enterprise security expectations.

There are zero hidden fees. All fees are clearly listed on pricing page.