Step: hypershift-launch-wait-for-nodes

This step blocks indefinitely until the specified nubmer of nodes have a `Ready` condition value of `True`. It assumes the `KUBECONFIG` environment variable is set as part of a workflow that creates clusters.

Container image used for this step: cli

cli resolves to an image built or imported by the ci-operator configuration (documentation).

Environment

In addition to the default environment, the step exposes the following:

Variable Name Type Variable Content
WAIT_FOR_NODES_COUNT Parameter[?] The number nodes which should become ready before proceeding. (default: 2)

Source Code

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
#!/bin/bash
set -euo pipefail

echo "$(date) Waiting for all nodes to join the cluster"
while true; do
  [[ $(oc get nodes --output go-template='{{ len .items }}') -eq "${WAIT_FOR_NODES_COUNT}" ]] && break
  sleep 5
done

echo "$(date) Waiting for all nodes to become ready"
oc wait nodes --all --for condition=Ready --timeout 24h

echo "$(date) All nodes are ready"

Properties

Property Value Description
Step timeout[?] 15m0s Limits the execution time of the step.
Termination grace period[?] 5s Period of time until SIGKILL signal is sent to the test pod (after SIGTERM signal is sent).
Resource requests (cpu) 100m Used in .resources.requests of the pod running this step.

GitHub Link:

https://github.com/openshift/release/blob/master/ci-operator/step-registry/hypershift/launch/wait-for-nodes/hypershift-launch-wait-for-nodes-ref.yaml

Owners:

Approvers:

Reviewers:

Source code for this page located on GitHub