Configuring access for Kubectl to private GKE cluster
A guide to how to configure access for Kubectl to private Google Kubernetes Engine cluster with different level of restricted access
Spinnaker stores in redis it’s pipeline execution history. The default configuration deploys a pod in the kubernetes cluster. If the pods gets restarted, all of the history gets lost.
Using ElastiCache removes this possibility.
We are going to use a terraform script to create all that we need for this. The script is located in this repository: https://github.com/armory/terraform/tree/master/elasticache
This script creates an ElastiCache cluster using a Redis engine, it’s based on an already existing vpc using its subnet ids, this to allow connectivity to the Spinnaker cluster.
In the file “terraform.tfvars” you need to set some variables.
-subnet-1-cidr: the eks subnet cluster -subnet-2-cidr = the eks subnet cluster -availability_zones: zones where a node of the cluster will be available. This needs to be equal or less than the number of nodes in the cluster -security_group_ids: The eks nodes sg
terraform init terraform plan -var-file=terraform.tfvars terraform apply -var-file=terraform.tfvars
After run the script the infrastructure should be created and the output of the script is the primary endpoint of the ElastiCache cluster.
To validate Spinnaker has connection to elasticache you need to exec into gate pod and run
nc -zv <endpoint from output (elasticache primary endpoint in the aws console)> 6379
you should see something like:
bash-4.4$ nc -zv myelasticache.usw2.cache.amazonaws.com 6379 myelasticache.usw2.cache.amazonaws.com (0.0.0.0:6379) open
If it’s not open, you should verify the subnet ids, and security groups to validate it is correct.
To configure spinnaker to use this cluster follow:
Using elasticache in aws eks cluster: Run terraformer scripts: Need to output the dns endpoint
If terraformer is enabled: add to terraformer-local.yml
redis: host: <elasticache-endpoint> port: <port>
If dinghy is enabled: add to dinghy-local.yml
services: redis: address: <elasticache-endpoint>
overrideBaseUrl: redis://<elasticache-endpoint>:<port> skipLifeCycleManagement: true
Disable automatic Redis configuration in Gate by adding the following to your gate-local.yml file:
redis: configuration: secure: true
To delete the default redis pod run:
kubectl -n spinnaker delete deployment spin-redis