1

While deploying an instance of screwdriver.cd via helm chart, where do we specify how much CPU every job would consume?

Every new build triggered currently via screwdriver spins up a pod consumes which request for below resources:

"resources": {
    "limits": {
        "cpu": "2",
        "memory": "2Gi"
    },
    "requests": {
        "cpu": "2",
        "memory": "2Gi"
    }
}

Where can we update/overwrite this in helm chart while deploying it? I don't want every build to consume 2 CPUs of my cluster.

I did update environment variables in the pipeline to use less CPU as instructed here:

shared:
    environment:
        CPU: micro
        MEMORY: micro
Tannu Priya
  • 313
  • 2
  • 15

1 Answers1

2

For Helm chart, since builds run on LOW cpu and memory by default, you can configure the default LOW values under the queue section in the values.yaml file.

To use less CPU and MEMORY in your Screwdriver pipeline, you'll need to use screwdriver.cd/cpu and screwdriver.cd/ram annotations in your screwdriver.yaml file.

For example, to use micro CPU and MEMORY, you should have something like this:

shared:
    annotations:
        screwdriver.cd/cpu: MICRO
        screwdriver.cd/ram: MICRO
Tiffany
  • 86
  • 3