Summary
The user is attempting to upgrade Flyte in their GCP deployment using Terraform but is facing issues with Helm not reading variables from their values file. They have run helm repo update and terraform plan, and are weighing the pros and cons of hardcoding a version versus using an input variable or the latest version from the repository cache. The terraform apply process takes 6 minutes and 30 seconds before hitting a client rate limiter error, which they suspect is due to a network issue. They also encounter a context deadline exceeded error when some Flyte Pods fail to start, resulting in a Terraform timeout. Everything is functioning except for one webhook pod, which has an autoscaling issue. The user suggests adding a specific version line in their local copy of the Terraform file for a more reliable upgrade.