Key Features
- Simple ratio-based system for allocating memory and compute
- High-availability groups to ensure instances have hypervisor isolation
- Hyperscale instances shared across hypervisors for single giant instances
- Live-updating context system for updating networking and variables passed to instances
- Flexible boot options for network/PXE, ISO images, and disks
- SSH key management and access groups
Case Study: Joviam
"HyperCloud's ratio-based approach allowed us to define a standard VM type which is easy to understand. Customers love that they can move a simple slider to get the exact number of cores they need."
~ Ruben S, engineer at Joviam
Simple resource allocation
The industry standard for cloud platforms are arbitrarily-defined instance types with set memory and virtual CPU cores, which isn't that far removed from how virtual private servers (VPSs) were sold. Named for t-shirt sizes or other nomenclature, the myriad of options can be overwhelming, confusing, and difficult to scale.
HyperCloud scales resources based on simple ratios. As the amount of allocated memory increases, so too does the vCPU core count. For example, an administrator may define a "standard" cloud instance as having 1 vCPU core per 4GiB of memory.
These ratios can be defined into multiple instance classes specific to the ratios workloads require, and can be selectively enabled and disable for specific organisations under a reseller tree.
High availability groups
HyperCloud instances can be managed through the use of High Availability groups (Hagroups). When two or more instances are grouped, they're guaranteed to boot on disparate hypervisors.
These are perfect for databases or load balancers with multiple replicas or workers, as they're isolated from hardware and local network failure. This renders any architecture built on the platform considerably more robust.
Live-updating contextualization system
Hypercloud instances have a set of context variables supplied by default. These generally include system information such as IP and gateway addresses, but they can also include user provided information.
Upon information changing by the end user or platform, the context values inside the vm are updated. That update event can be scripted against by the user, allowing for easy automation and live contextualization.
Flexibility
Instances can be provisioned with defined ratios for vCPU cores and memory, SSH keys for Linux and FreeBSD, and can be easily scaled up and down as workloads require. In this example, the HyperCloud install uses High Memory, Standard, and High CPU for different resource ratios.
Flexible boot options
Many large scale environments have found that PXE based deployment is an easy way to automate against large amounts of real hardware. HyperCloud provides a PXE boot based environment based on iPXE, a commonly used PXE ROM, to allow that deployment infrastructure to work seamlessly between the existing hardware, and new cloud instances.
HyperCloud also allow for disks to be marked as CDs, which then allows you to boot instances from those CDs. This is useful for creating images.
