An engineer needs to set up a new environment for processing non-critical batch jobs on Google Kubernetes Engine (GKE) while ensuring efficient resource utilization and enabling direct administration from their local command line interface. Which two steps should the engineer execute to achieve this setup?
- Run gcloud container clusters get-credentials to fetch authentication credentials and update the local kubeconfig file.Answer
- Provision a Spot node pool within the GKE cluster to run stateless, fault-tolerant batch processing workloads at reduced compute costs.Answer
- CRun gcloud config set container/cluster to authenticate the CLI and establish the active kubectl context for cluster management.
- DDeploy GKE Autopilot clusters when custom Linux kernel parameter modifications and root access to individual node operating systems are required.
Answer
The engineer must run 'gcloud container clusters get-credentials' to populate the local kubeconfig for kubectl access, and provision a Spot node pool to cost-effectively run stateless batch processing workloads.
To interact with a GKE cluster via kubectl, the command 'gcloud container clusters get-credentials' must be used to generate the correct cluster context entry inside kubeconfig. Additionally, running stateless batch workloads on Spot node pools provides significant cost savings while tolerating potential node preemption.
Step-by-Step Solution
Key Concept
Configuring GKE cluster credentials and selecting appropriate workload compute tiers