Glasswings proprietary VPA Automation continuously optimizes your Kubernetes environment by autonomously adjusting workloads' CPU and memory to optimal values, while respecting any predefined settings customers may configure (e.g. Keeping QoS).
Built for dynamic, production-grade environments, our automation engine works seamlessly alongside Horizontal Pod Autoscalers (HPA) to maintain workload stability, performance, and cost efficiency. Glasswing automations safely adjust resource allocations without requiring manual oversight, helping your teams focus on innovation instead of infrastructure management.
When Glasswing identifies resource waste and its recommendations are considered significant according to the policy assigned to a workload, Glasswing automation will gradually reduce resources. The system carefully monitors the proper workload functioning after each resource-decreasing step to ensure the best availability.
When workloads consistently exceed their allocated resources, Glasswing’s automation proactively increases resource allocations to maintain application health and prevent performance degradation. This ensures mission-critical services remain stable and responsive, even under changing load conditions.
To ensure Glasswing optimizes workloads appropriately, it’s essential to assign the desired Optimization Policy to each workload.
The optimization policies allow you to specify how your resources should be allocated in order to support the individual needs of your workloads. Assign the policies that best suit your environment and business goals, depending on whether you want to maximize cost savings or provide extra headroom to maintain the resilience of mission-critical services
Detailed information on policies and how to create and assign them can be found in under Rightsizing Policies page
While enabling automation can resolve performance degradation and instability, it’s important to review workloads with existing health issues (e.g., frequent restarts or out-of-memory events). Automating unhealthy workloads that are not under-provisioned, may mask deeper issues unrelated to resource sizing. Glasswing provides an alert in the workload drawer tab to indicate all identified health issues as well as a filter to view all unhealthy workloads.
Glasswing is fully aware of your Horizontal Pod Autoscaler (HPA) settings when applying automated changes. Because we always prioritise resilience over cost, Glasswing automatically applies a cooling-off period after each optimisation on HPA-controlled workloads. As a result, these workloads may achieve lower cost savings than similar workloads that are not managed by HPA.
When Glasswing detects resiliency issues—such as OOM kills or CPU throttling—it automatically scales resources up to restore stability and protect overall service health.
