dev 2 dev

CAST AI created by developers for developers

K8s Kubernetes Service

Kubernetes Service

Get a no-nonsense, vanilla k8s cluster ready to use in just a few minutes, on one or more cloud platforms. Break free from cloud provider’s specific APIs, concepts, and limitations. The cluster will always follow the configuration of your choice. We harden your deployment with features like private nodes, a just-in-time firewall, and automated TLS provisioning.

Infrastructure as Code

Use our industry-standard Terraform Provider, CAST APIs or, Command Line Interface (CLI) to integrate CAST operations with your CI/CD pipelines.

Virtual Multi-Cloud Private Network

This serves as the underlying secure communication overlay - all pods have immediate access to each other and native cloud services over private IP.

Outstanding observability

We use Cilium to enhance network performance, security, and cluster observability. Benefit from logging with ElasticSearch and Kibana, built-in metrics from Prometheus and Grafana.

Cloud Cost Optimization

Cloud Cost Optimization

Without clear visibility, it’s difficult to understand if you’re running the most optimized version of your infrastructure and what potential steps you can take to improve it. CAST AI brings resource and cost optimization to Kubernetes in one seamless platform across all clouds. Our automated scaling features and optimization engine automatically apply cluster changes based on changing workload conditions.

Cost Savings Price Arbitrage

The CAST AI optimization algorithm is always looking for opportunities to replace VM Nodes with less expensive alternatives that do the job.

Node Replacement

Based on your usage, CAST AI consolidates or replaces nodes in your cluster. We move all the workloads automatically to avoid downtime and make sure to remove older notes once they’re safely vacated.

Node Movement

Sometimes, you can achieve a better cost vs. performance balance on an adjacent cloud service. CAST AI can move your cluster resources around easily, following the policies you set for it.

Intelligent Scheduling

If you don’t use your clusters all the time, CAST AI allows pausing them without losing configuration or state. Once you need a cluster again, you’ll find it ready to go.

Disaster Recovery

Disaster Recovery & High Availability

It’s impossible to avoid system faults altogether, but just because a component fails it doesn’t mean that it should cause you downtime. CAST AI is designed to make workloads highly resilient to failure, even if the entire cloud service goes down. By deploying one logical cluster across multiple clouds or cloud regions, we fortify your cluster deployment and protect it from faults if that would cause availability and performance issues.

High Availability Components

CAST AI can provision your cluster with highly available components such as multiple load balancers, VPN connections, master nodes with a replicated etcd database, all the way to distributed databases.

Automatic Traffic Removal

CAST AI has partnered with NS1 to provide a globally load-balanced CNAME to your cluster. It automatically fails over and removes traffic from a failed cloud if it’s experiencing issues. Once it detects a failure, CAST AI fails over within 60 seconds (but we’re working hard to lower that threshold).

Support for Disaster Recovery and High Availability

CAST AI supports two modes for disaster recovery and high availability:

  • Active-Active-(Active) - allowing two or more clouds to cooperate in a single application scope, even when it requires a persistence layer such as a relational database.
  • Active-Passive (Hot Standby) - here, disaster recovery relies on the primary cloud provider to serve the customer application. The provided DNS CNAME detects a cloud outage and automatically redirects traffic to the secondary standby cloud.

Scaling Workloads

Scaling workloads

CAST AI providers automated load balancer and TLS provisioning. You also get a comprehensive set of auto-scalers to grow your workload footprint automatically - and then shrink it to avoid wasting resources. Scaling happens by using the same mechanism, regardless of the cloud services you choose.

Pod Auto-Scaling

Configure and watch as pods automatically scale to meet the demands of your workload. CAST AI can scale pods horizontally (multiple replicas) based on Kubernetes Event Driven Autoscaling (KEDA).

Node Auto-Scaling

Define rules for how your infrastructure should scale in the Policy Engine. Set limits, CPU, and Memory headroom for growth and the types of activity that trigger cluster scaling. CAST AI will automatically select nodes that fit your workload - and then let these nodes go when you no longer need them.

Cloud Auto-Scaling

CAST automatically increases capacity where your workloads are. Use additional clouds to burst resources or distribute your workloads evenly across the cloud providers you’ve selected.

Automated scaling events are integrated with an audit logging feature to help you keep track of changes.

Empower your cloud with CAST AI kubernetes

Secure and managed Kubernetes service.
Deploy, operate, and cost optimize Kubernetes applications in any combination of clouds.