Skip to main content
Explore
Log in
Get started
Someone wanted to learn this too, so Grasp built them a personal learning path.
Create your own
GCP Kubernetes infrastructure scaling
Module 1
Foundations for Autoscaling Event-Driven Services
1
Competing Consumers for Pub/Sub Scaling
Describe the competing consumer pattern for scaling Pub/Sub subscribers by running multiple replicas.
Describe the competing consumer pattern for scaling Pub/Sub subscribers by running multiple replicas.
2
Optimizing Container Performance with Kubernetes Resources
Define Kubernetes resource requests and limits for a container to ensure predictable performance and scheduling.
Define Kubernetes resource requests and limits for a container to ensure predictable performance and scheduling.
3
Inspecting Pod Resources with `kubectl top` and `kubectl describe`
Use `kubectl top` and `kubectl describe` to inspect the real-time resource usage of a running pod.
Use `kubectl top` and `kubectl describe` to inspect the real-time resource usage of a running pod.
4
HPA: Metrics, Polling, and Scaling Loop
Explain the mechanism of the Horizontal Pod Autoscaler (HPA), including its metrics polling and scaling loop.
Explain the mechanism of the Horizontal Pod Autoscaler (HPA), including its metrics polling and scaling loop.
5
Scaling Queue-Based Workloads: Beyond CPU/Memory
Justify the need for custom metrics over standard CPU/memory for scaling queue-based workloads.
Justify the need for custom metrics over standard CPU/memory for scaling queue-based workloads.
Module 2
Implementing Custom Metric Autoscaling on GKE
6
Installing and Configuring the Custom Metrics Stackdriver Adapter
Install and configure the Custom Metrics Stackdriver Adapter to expose GCP metrics to the Kubernetes API server.
Install and configure the Custom Metrics Stackdriver Adapter to expose GCP metrics to the Kubernetes API server.
7
Autoscaling with Custom Pub/Sub Metrics
Write an HPA manifest that targets the `pubsub.googleapis.com|subscription|num_undelivered_messages` metric.
Write an HPA manifest that targets the `pubsub.googleapis.com|subscription|num_undelivered_messages` metric.
8
HPA Autoscaling with Pub/Sub Backlog
Deploy and test the HPA, verifying that replica counts adjust based on Pub/Sub backlog size.
Deploy and test the HPA, verifying that replica counts adjust based on Pub/Sub backlog size.
9
HPA and Cluster Autoscaler: Dynamic Scaling Explained
Explain the role of the Cluster Autoscaler in provisioning new nodes for pods scheduled by the HPA.
Explain the role of the Cluster Autoscaler in provisioning new nodes for pods scheduled by the HPA.
10
GKE Autoscaling: Node Pool Configuration
Enable and configure cluster autoscaling on a GKE node pool to handle increased pod counts.
Enable and configure cluster autoscaling on a GKE node pool to handle increased pod counts.