Introduction to Infrastructure
One of the core features of the Cycle Platform is global multi and hybrid-cloud infrastructure management. Cycle brings a new paradigm to cloud computing, abstracting servers from various cloud providers and on-prem/custom datacenters into a single cloud, managing them and keeping them up-to-date, while ensuring ownership and control of the servers remains firmly in the hands of the user.
The Cycle Infrastructure Paradigm
When using Cycle, infrastructure is deployed via hub integrations to one or more cloud providers, or to on-prem via the IAL. The cloud provider accounts connected via the integration are owned by the user, and Cycle deploys infrastructure via that account. Cycle (the company) does not own any customer infrastructure.
Once the desired integrations are configured, all deployments are done from Cycle. There is no need for someone to understand in-depth how to deploy servers via AWS EC2, GCP, etc - Cycle abstracts the complexity of infrastructure management into a single place, making it incredibly easy to deploy servers across a multitude of different providers.
Cycle can deploy and manage both bare-metal servers and VMs - all infrastructure deployed via hub integrations is treated exactly the same. The infrastructure ends up being one global pool of resources to run containerized workloads.
How Server Management Works on Cycle
When a server is deployed from Cycle via a hub integration, the server installs and boots CycleOS via iPXE - a thin operating system based on Alpine Linux . CycleOS is installed to RAM, and the RAM drive is set to readonly. This is done for security hardening purposes, and for the added benefit that, every time the server is restarted, it has the latest up-to-date version of the operating system, including any security patches and bug fixes.
By utilizing the same underlying OS for all servers, Cycle standardizes how everything across the platform works - eliminating massive amounts of complexity. This keeps things simple, with less moving parts as well. CycleOS essentially only knows how to provision networks and storage, run containers, and communicate with the core.
Cycle Compute Software
On top of the OS, Cycle runs just a few services to facilitate running containers. This software is automatically updated by our team, with new releases going out on average every 2 weeks. No server restart is necessary when these pieces of software are updated - it is all seamless behind the scenes.
The two main pieces of software are:
- agent: responsible for communicating with Cycle
- compute: responsible for actually running containers - Cycle's container runtime.
Benefits of the Cycle Model
Cycle's approach to infrastructure management keeps control firmly in the hands of the user, while eliminating much of the complexity involved with administrating a cloud. Multi-cloud deployments were, prior to Cycle, reserved only for the largest institutions and dev teams. By adopting the Cycle platform, even small teams can deploy and manage a hybrid cloud, utilizing compute where it makes sense, without being chained to rising costs of infrastructure simply because migrating costs more (check out how easy it is to move container instances between clouds on Cycle).
Get Started Deploying Infrastructure on Cycle
Follow one of these guides to integrate with a cloud provider and begin deploying infrastructure on Cycle: