Improve Digital Ocean guide

[adrianeriksen]

I ran trough the Digital Ocean guide and faced some problems. I got the cluster up and running after som troubleshooting, and I want to share how I did it.

The changes introduced with this PR are:

1. Updated selectors so the load balancer can reach the juice-balancer pod.
2. Added a HTTP listener to enable HTTP to HTTPS redirect.
3. Added some instructions on how to troubleshoot when the juice-balancer pod hangs on pending.

Problem 1: Provisioning the cluster

When installing multi-juicer on the cluster the juice-blancer pod is stuck on pending when I run kubectl get pods. I got the following message when I described the pod:

Warning  FailedScheduling   7m18s                default-scheduler   0/3 nodes are available: 3 Insufficient cpu. preemption: 0/3 nodes are available: 3 No preemption victims found for incoming pod..

The default node size is s-1vcpu-2g. I worked when I upgraded to s-2vcpu-2g. The downside is a more expensive cluster, approximately $66 per month.

Problem 2:

When provisioning the load balancer in step 4, I got this warning.

Warning  SyncLoadBalancerFailed  4m4s                service-controller  Error syncing load balancer: failed to ensure load balancer: load-balancer is not yet active (current status: new)

The provisioned load balancer can't reach juice-balancer. I get the following labels when I describe the juice-balancer pod.

Labels:       app.kubernetes.io/component=load-balancer
              app.kubernetes.io/instance=juice-balancer-multi-juicer
              app.kubernetes.io/managed-by=Helm
              app.kubernetes.io/name=juice-balancer
              app.kubernetes.io/part-of=multi-juicer
              app.kubernetes.io/version=5.3.0
              pod-template-hash=b756f7b9c

When I updated the selector, the load balancer provisioned as expected.

[J12934]

Ahhh yes 👍
I've recently changed the labels to be more consistent, forgot to update the guide(s).
Thank you for the fix and the feedback 🙌