Use Deployment objects instead of Pod to manage jenkins instances lifecycle

[akram]

Currently the jenkins operator monitors pod configuration changes and performs restarts of pods in case of a change. Also, it monitors the existence of the pod and recreates it if it deleted.
Kubernetes provides Deployment object to perform these 2 features [1].

It would be preferable if the jenkins-operator instantiates a Deployment object and lets the kubernetes Deployment Controller manage the pod lifecycle.

[1] https://kubernetes.io/docs/concepts/workloads/controllers/deployment/

[tomaszsek]

Pod gives full control of the Jenkins. I have an idea to spin up two pods at the same time and switch traffic from one to another when the old one needs to be killed. With deployment you can't do this.

[pawelprazak]

I wonder if we could leverage the deployment controller and still do the blue-green deployment

[akram]

Rolling-update allows indeed to achieve blue/green deployment.
Deployment are backed with ReplicaSet which are controlled by replication controller.
And the replication controller supports the rolling update feature [1]
We have to pay attention at 2 things:

• the volume type / access mode which must support RWX mode (Read Write by multiple pods at a time)
• define properly liveness and readinessprobe which determines the triggers to switch traffic to the new pod

https://kubernetes.io/docs/tasks/run-application/rolling-update-replication-controller/

[waveywaves]

@tomaszsek can you assign this to me ?

[triesz]

hi @tomaszsek, @waveywaves et al, sorry, I am slightly late on the show, but would it not be more beneficial to use StatefulSet instead of Deployment? I.e. volumeClaimTemplates would allow stable storage...
https://kubernetes.io/docs/concepts/workloads/controllers/statefulset/#stable-storage

[snooyen]

@tomaszsek @waveywaves

Any movement on this? Is there any way I can help move this along?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If this issue is still affecting you, just comment with any updates and we'll keep it open. Thank you for your contributions.