Hello fellow contributors!
About 10 days ago a Digital Ocean Kubernetes cluster has been added to the available agents for ci.jenkins.io build jobs. (More details here: jenkins-infra/helpdesk#2651)
For the more curious, you can check on which kubernetes
cluster your jobs are running by consulting the full console output in
Jenkins UI.
In the first lines of the logs there should be a section describing the
pod created. We've added the cluster name in its metadata labels, ex:
ci.jenkins.io/job/Core/job/jenkins/job/PR-6317/1/consoleFull
17:35:20 Agent jnlp-maven-11-bh99z is provisioned from template jnlp-maven-11 17:35:20 --- 17:35:20 apiVersion: "v1" 17:35:20 kind: "Pod" 17:35:20 metadata: 17:35:20 labels: 17:35:20 jenkins: "slave" 17:35:20 jenkins/label-digest: "25a08295458738d4af47a3e71f274bebae57766b" 17:35:20 jenkins/label: "container_kubernetes_doks_maven-11_jdk11" 17:35:20 name: "jnlp-maven-11-bh99z"
We can see here "doks" in "jenkins/label",
meaning this job is running on Digital Ocean.
If it were running on AWS, it would have been "cik8s" instead.
The specs of this cluster nodes are similar to the
existing ones on AWS, there shouldn't be any major changes, except
decreasing our AWS billing notes.
Have you seen or felt a difference in your builds in terms of execution time or error rates since?
Please don't hesitate to let us know if you encountered any problem or if you have any remark and/or question.
Kind regards,
Hervé Le Meur for the Jenkins Infra Team