OvrClk is a serverless platform provider that delivers high-performance workloads at the edge by optimizing for low latency and high efficiency
OvrClk provides functional and application abstractions over a self-healing fabric of container-native infrastructure by leveraging modern cloud-native ecosystem and eliminates the need to install, operate, and scale a high-performance cluster management solutionGet Started Request Invite
OvrClk provides a federated network spanning across datacenters and cloud providers by relying on Linux containers that are connected to an extremely secure global Virtual Private Network and backed by a dedicated high-speed network backbone designed to improve consistency guarantees while optimizing for policy-aware data deployments across the globe
With an industry average of just 15% per machine utilization, OvrClk is designed to significantly increase the utilization by up to 80%.
It does so by algorithmically managing placement of applications on the system to efficiently allocate jobs by maximizing system resource usage. At any given time, it ensures applications has the exact amount of resources it requires without overhead.
OvrClk makes infrastructure fault tolerant to erosions and natural disasters. When a fault is detected, the system ensures the an instance of the application is always running.
Faults are inevitable. Instead of just focussing on reducing faults, the system focusses on recovering quickly. This is achieved by decoupling build and run time of application that result in immutable images. OvrClk is a distributed state machine. At any given time, it maintains several copies of the state spread through out the globe.
OvrClk provides primitives to build disposable and mutable infrastructures. These controls allow launching and disposing clusters around the globe seamlessly using an intuitive command line interface.
It provides for reliable and frequent container image build and deployment with quick and easy rollbacks (due to image immutability) using an simple git based deployment.
The intrinsic security of containers, as implemented by kernel namespaces and cgroups and the “hardening” security features of the kernel provide a extremely secure runtime for container applications.
The machines run a hardened minimal operating system optimized designed to securely run containers. The machines in a cluster communicate with each other in an encrypted overlay network using UDP to encapsulate IP packets. The cluster are backed by a TLS based VPN. The encryption keys are centrally managed in secure vault that seal automatically when under a threat.
OvrClk is built for hyper, warehouse and planetary scale.
Launching a cluster usually takes under 10 minutes. Run thousands of containers in seconds. The system is build on years of experience running production scale system at Google
A simple and intutitive utility to manage all aspects of clusters and applications on OvrClk
Powerful and easy to use RPC and REST API to extend OvrClk
Simplicity of PaaS with the flexibility of IaaS and enables you to run much more than just 12-factor apps
Powerful controls to manage and respond to sudden increases in application load automatically
Auto-placement, auto-restart, auto-replication using immutable application images to recover from faults due to errosions and unknown disasters with strong backup facilities
Dynamic DNS for auto discovering services and applications inside and outside of the cluster
Securely connect OvrClk to existing infrastrcuture
Single integrated global network backed by a TLS based OpenVPN
Instant scaling to thousands of instances across the globe
Secure vault to manage sensitive information such as API keys, passwords and certificates with Leasing and Renewal
Rapid debugging and analysis to see application performance and logs
Git-based atomic application deployments with powerful controls of orchestrating powerful deployment workflows
Out of the box buildpacks to instantly build and run applications go, ruby, node.js, java and many more
Powerful discovery, build and runtime support for highly decoupled and focussed services with isolation controls for multiple tenants
Containers running on a single machine all share the same operating system kernel so they start instantly and make more efficient use of RAM
The system is build on years of experience running production scale system at Google