Google has unveiled a new ‘air-gapped’ appliance as part of its Distributed Cloud offering, pledging to bring cloud and AI capabilities to the tactical edge where internet access is missing.
The company explained that the appliance solution runs Google’s cloud infrastructure stack, including Kubernetes clusters, data security services and the Vertex AI platform to essentially create a cloud-in-a-box, ‘air-gapped’ – rendering it physically segregated and incapable of establishing an external connection – either wirelessly or physically.
Why the complete isolation? After all, the major benefit of operating in a public cloud is to grant your applications access to the masses of data, functionality and partner options available from that position. However, an outside cloud connection would also make it and its contents and connections insecure and vulnerable, and so it is typically ruled out for many secure or sensitive applications.
Military, government and banking installations, for instance, are simply barred from running on a public cloud for broad security reasons. Not just to prevent a bad actor getting access to or hijacking data, but to diminish the possibility that some sort of cloud outage might ripple through and knock over the system – a danger concentrating IT minds in the wake of the recent global CrowdStrike outage.
Ordinarily, organisations with mission-critical workloads lack access to important cloud and AI capabilities when in demanding edge environments. In this case, Google claims its air-gapped solution has security clearance for most US military or Department of Defense (DoD) use cases, having received DoD Impact Level 5 accreditation for its security features. This includes data encryption, data isolation, firewalls and a secure boot. Developers can use the appliance’s computing, storage and networking components as well as the software running on top via the same APIs used in Google Cloud, the company claims.
Google maintains the solution will appeal in disconnected or mobile situations, such as disaster zones where the internet has been knocked out, long-haul trucking operations or remote research stations. The integrated hardware and software solution unlocks real-time local data processing for AI use cases, such as object detection, medical imaging analysis and predictive maintenance for critical infrastructure.
The appliance, housed in a rugged case and weighing about 45kg, may be carried into location on foot by two people.
– Ian Scales, Managing Editor, TelecomTV