The world is becoming increasingly digitized, with more devices and more data being deployed and used at the edge. The conventional approach of sending edge data to the cloud for processing is often impractical due to network limitations and costs. Instead, a new approach is needed, one that brings the cloud operating model to the edge, empowering organizations to run applications in distributed environments.
The Challenges of Edge Orchestration and Management
Traditional IT Assumptions vs. Edge Reality
Traditional IT systems are built on assumptions that don’t hold true at the edge. These systems expect consistent high-speed networks, abundant computing resources, uniform server configurations, and easy scalability. However, edge computing environments are fundamentally different. They operate across diverse locations with unreliable network connections, limited computing power and storage, and challenging environmental conditions that affect power and cooling capabilities.
Automation Requirements
The distributed nature of edge computing creates two critical automation needs. First, systems must support zero-touch provisioning due to the scale and geographic spread of deployments. Second, application deployment must be fully automated to handle diverse devices and applications without manual intervention. These automated systems must also manage the full application lifecycle, including updates, patches, and platform changes.
Management Platform Needs
Traditional management approaches using single-purpose appliances and multiple disconnected tools aren’t sufficient for edge computing. Organizations instead need:
- A centralized management platform that provides unified visibility across all edge devices.
- Comprehensive remote monitoring capabilities with logging and alert systems.
- Flexibility to handle diverse applications, from AI inference to streaming and security tools.
- Scalability to manage thousands of edge devices without becoming overwhelmed by complexity or costs.
Zededa Can Help
Zededa is a company that aims to solve these challenges by bringing the cloud operating model to the edge. Zededa designed its platform to manage and orchestrate applications across thousands of edge devices, providing a secure and scalable solution. At the heart of Zededa’s platform is EVE (Edge Virtualization Engine), an open-source operating system designed specifically for edge environments. EVE is a lightweight, secure, and pluggable OS that provides a foundation for running a wide variety of workloads. It was open-sourced into the Linux Foundation Edge group and is now Apache licensed and community developed.
Zededa Architecture
The Zededa platform is an API-only solution, meaning that there is no local console or SSH access to the devices. This ensures a higher level of security and makes configuration management more consistent and centralized. The architecture is composed of:
- Immutable OS: The root file system is read-only, which prevents tampering and ensures the integrity of the operating system.
- Embedded Hypervisor: The platform supports multiple workloads, including virtualized network functions, legacy virtual machines, and containerized workloads. KVM is the default hypervisor, and Acorn and Zen are also supported.
- IO Virtualization: The platform provides extensive support for IO virtualization, allowing for the passing of USB connections, camera feeds, and GPUs to applications.
- Pluggable Architecture: Zededa can be built on different kernels, with Alpine Linux as the default. Red Hat and SUSE are also supported.
- Zero-Touch Updates: Zededa supports zero-touch updates, using a dual-partition environment that allows for seamless updates and rollback in case of failure.
- Measured Boot: Zededa uses measured boot, which measures PCR values of the device and software, and stores them in the TPM. The cloud compares this to the stored checksum, and if they match, it allows the applications to run; otherwise, the device is placed in an offline state.
- Connection to Zededa Cloud: The platform connects to the Zededa Cloud Controller, which is a cloud-based SaaS platform that manages all the devices. The cloud controller is an API controller where policies and configurations are set.
Features and Benefits
The combination of Zededa and EVE offers numerous features and benefits for managing and orchestrating edge applications, including:
- Scalability: Zededa designed the platform to handle large deployments (in excess of 170,000 nodes). Customers can manage their deployments as a fleet, applying configurations and updates across all devices with a single click.
- Security: Zededa’s platform has a strong security focus, with features like measured boot, encryption at rest and in flight, read-only file systems, and zero-trust architecture.
- Flexibility: The platform is software and hardware agnostic, supporting x86, Arm, and RISC architectures. It also supports various workloads, including virtual machines, containers, and Kubernetes.
- Connectivity: The platform can manage diverse connectivity options, including 2G, 4G, 5G, cable, satellite, and microwave. It includes policy engines that detect the type of connectivity and set the primary and backup routes and multipath connections.
- Remote Management: The platform allows for remote management and operations of edge devices, eliminating the need for local IT staff. It includes a cloud controller with an interface for logging, monitoring, and alerts, allowing for real-time support.
- Zero-Touch Provisioning: Organizations can deploy devices with a zero-touch process: devices are simply plugged in, and they connect over the available network, download their configuration, and deploy the necessary applications.
- Application Marketplace: Zededa offers a global marketplace with commercial and open-source software solutions. Customers can also create a private marketplace for their own applications and legacy systems.
- Infrastructure as Code: Zededa pushes infrastructure as code as well as the app marketplace to the edge to be deployed, allowing for eventual consistency in all devices in the fleet.
Limitations
While Zededa’s platform offers numerous benefits, there are some potential limitations to consider:
- Complexity: Managing large fleets of edge devices with diverse connectivity options can be complex and requires careful planning and implementation. While Zededa offers automation through an API, some organizations may find it difficult to integrate into current systems.
- Orchestration Limitations: Some customers have found that their orchestration layer is not keeping up with the load for Kubernetes, and Zededa is working to address this issue. The current solution is sharding, but this is not ideal. Zededa is developing a more scalable orchestration layer.
- Data Protection: Because most deployments are single node, data is backed up and restored rather than using techniques like snapshots. In stateful applications that are long-term and have large data sets, there may be a need to use multiple nodes to prevent data loss.
Why This Matters
Edge computing presents both a massive opportunity and a critical challenge. While organizations can now collect and process data closer to its source across countless distributed devices, they need sophisticated platforms designed specifically for the complex demands of edge environments – from remote deployment to seamless scaling.
Zededa provides a solution that enables organizations to bring cloud-like agility, scale, and security to their edge deployments. By pushing cloud capabilities to the edge, Zededa enables businesses to make real-time decisions locally, reducing latency and costs. The ability to manage fleets of devices remotely, in a standardized and secure manner, enables organizations to focus on their core business rather than being bogged down by the complexity of edge computing.
If you need a cloud-like operating model at the edge that is secure, scalable, flexible, and easy to manage or you’re facing the challenges of deploying, orchestrating, and managing applications at the edge, especially with large-scale, geographically dispersed deployments, you should consider if ZEDEDA can help.