Google today took the wraps off an appliance edition of an air-gapped instance of the Google Distributed Cloud platform based on Kubernetes clusters.
Sachin Gupta, vice president and general manager for Google Cloud, said this appliance extends the portfolio of options organizations have today for deploying Google Distributed Cloud in environments where connections to the internet are either not feasible or allowed.
For example, organizations looking to deploy Kubernetes clusters in an environment that required a ruggedized appliance can now use an instance of Google Distributed Cloud that is for all intents and purposes of a cloud computing environment in a box, he added. The approximately 100-lb appliance has achieved Impact Level 5 accreditation, the highest level of security controls and protection required for unclassified, but sensitive information defined by the U.S. government. It meets stringent accreditation requirements like MIL-STD-810H, ensuring reliable operation even in challenging scenarios. Those specifications make it feasible, for instance, to set up a complete IT environment in the wake of a natural disaster.
That approach makes it possible for developers to build container applications or monolithic applications deployed on an instance of open-source KubeVirt software to encapsulate virtual machines using a familiar set of interfaces, noted Gupta.
IT operations teams could then either manage Google Distributed Cloud in isolation or remotely connect to the platform via a private network. Initially, the platform is configured by downloading an instance of Google Distributed Cloud via a secure connection to Google Cloud Platform (GCP).
Kubernetes in Air-Gapped Environments
The number of use cases involving Kubernetes in air-gapped environments continues to expand as more organizations look to process and analyze sensitive data at the point where it is created and consumed. The appliance edition of Google Distributed Cloud makes it possible to bring compute resources to where that sensitive data resides using a platform with built-in encryption, data isolation, firewalls and secure boot capabilities.
In addition, organizations are also starting to build and deploy artificial intelligence (AI) models that access local data to apply, for example, speech and optical character recognition or make use of generative artificial intelligence (AI) to make it simpler to decipher manuals, noted Gupta.
Each IT organization should determine what level of isolation might be required for any given use case. In some cases, internet connectivity that enables applications to take advantage of elastic cloud services is still going to be required. There is also a rack-mounted version of Google Distributed Cloud designed to be deployed in less rugged IT environments.
The one certain thing is that as IT continues to evolve there will soon come a day when there are more computing resources at the network edge than there are in the cloud or a local data center. The challenge will be finding a way to centrally manage all those highly distributed compute resources in a world where there are simply not enough IT professionals to physically manage every platform deployed either at the network edge or in an air-gapped IT environment.
Related