logo

Use Case No.02

Automated Resource Scaling for Critical Workloads in Cloud

Use Case:
Event: High CPU/GPU usage, memory consumption, or storage capacity thresholds reached during access of application or inferencing of AI models.
Problem: Large scale end-user facing workloads are often resource-intensive and require dynamic scaling of compute, memory, and storage resources in cloud to meet varying demands.
Workflow:

Trigger Scaling Workflow

The event (resource consumption threshold) triggers an automatic resource scaling workflow.

Scale Down Resources

Once the workload is reduced, the workflow scales down resources to optimize costs.

Monitor Usage Metrics

The system continuously monitors usage metrics and sends notifications to administrators in case of resource bottlenecks.

Allocate Compute Resources

Workflow images are triggered to allocate additional compute resources (e.g., scaling virtual machines or containers in the cloud) based on demand.

Remove Scaled Down Resources

Remove the scaled down resources back to a Load Balancer of service discovery framework.

Add Scaled Up Resources

Add the scaled up resources back to a Load Balancer of service discovery framework.

Outcome: Automated resource scaling ensures optimal use of cloud infrastructure, preventing performance bottlenecks during AI/ML training or inference workloads as well as large user-facing applications.