Single-region insecure cluster on EKS tutorial

[jseldess]

How to run insecure on EKS using configs and helm.

Running secure will be handled in a separate PR.

• Secure with standard configs doesn't work due to some issue with Kubernetes-created CSRs on EKS. Will need to use different configs to use certs generated outside of Kubernetes.
• Secure with helm won't work because it uses standard configs. We will probably need distinct helm charts to GKE and EKS. Opened this issue: Helm charts for secure CockroachDB on EKS cockroach#38847. Secure with helm on EKS is therefore out-of-scope for now.

Addresses #5042.

[cockroach-teamcity]

This change is 

[cockroach-teamcity]

Online preview: http://cockroach-docs-review.s3-website-us-east-1.amazonaws.com/83f22f6949bc846dd6f8a3fa80c1ef8e4389eb4a/

Edited pages:

• v19.1/orchestrate-cockroachdb-with-kubernetes-insecure.md

[cockroach-teamcity]

Online preview: http://cockroach-docs-review.s3-website-us-east-1.amazonaws.com/f8b183d45cb6a95cccf12d0b5dab7483eeb4815e/

Edited pages:

• v19.1/orchestrate-cockroachdb-with-kubernetes-insecure.md

[cockroach-teamcity]

Online preview: http://cockroach-docs-review.s3-website-us-east-1.amazonaws.com/a893bf50c1e5c20a582e33f2c7aa726b296cb50b/

Edited pages:

• v19.1/orchestrate-cockroachdb-with-kubernetes-insecure.md

[jseldess]

A few issues found during testing that are probably independent of these instructions:

When I started a cluster via helm on EKs, the cockroach version was 2.1.1:

~$ kubectl run cockroachdb -it --image=cockroachdb/cockroach:v19.1.2 --rm --restart=Never -- node status --insecure --host=my-release-cockroachdb-public
  id |                                     address                                     | build  |            started_at            |            updated_at            | is_available | is_live
+----+---------------------------------------------------------------------------------+--------+----------------------------------+----------------------------------+--------------+---------+
   1 | my-release-cockroachdb-0.my-release-cockroachdb.default.svc.cluster.local:26257 | v2.1.1 | 2019-07-12 17:33:24.152488+00:00 | 2019-07-12 17:42:28.706545+00:00 | true         | true
   2 | my-release-cockroachdb-1.my-release-cockroachdb.default.svc.cluster.local:26257 | v2.1.1 | 2019-07-12 17:33:27.48001+00:00  | 2019-07-12 17:42:27.548286+00:00 | true         | true
   3 | my-release-cockroachdb-2.my-release-cockroachdb.default.svc.cluster.local:26257 | v2.1.1 | 2019-07-12 17:39:02.015971+00:00 | 2019-07-12 17:42:29.033669+00:00 | true         | true
(3 rows)

Don't understand why that would be given the chart says 19.1.2: https://github.com/helm/charts/blob/master/stable/cockroachdb/values.yaml#L8

Also, the helm config doesn't seem to have the app label, which we rely on to delete resources with one fell swoop (and maybe for other reasons). In the absence of that, I'll probably have to update the "Stop the cluster" section for helm.

[DuskEagle]

I don't see a reason why we couldn't add an app label to our Helm charts. It seems a lot of other Helm charts either have an app or a app.kubernetes.io/name label.

[cockroach-teamcity]

Online preview: http://cockroach-docs-review.s3-website-us-east-1.amazonaws.com/d702e9c706c7eff4537935c86ec66fc6ed7d5704/

Edited pages:

• v19.1/orchestrate-cockroachdb-with-kubernetes-insecure.md
• v19.2/orchestrate-cockroachdb-with-kubernetes-insecure.md

[DuskEagle]

I think the v2.1.1 happened because you didn't run helm repo update before starting. Could you add that to the instructions?

[DuskEagle]

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @DuskEagle and @jseldess)

---

v19.1/orchestrate-cockroachdb-with-kubernetes-insecure.md, line 144 at r1 (raw file):

    {% include copy-clipboard.html %}
    ~~~ shell
    $ helm delete my-release-cockroachdb

Adding a --purge flag here will free up the "cockroachdb" name in Helm in case a user wants to launch another CRDB cluster using Helm into the same Kubernetes cluster.

[jseldess]

I added a step for helm repo update as well.

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @DuskEagle)

---

v19.1/orchestrate-cockroachdb-with-kubernetes-insecure.md, line 144 at r1 (raw file):

Previously, DuskEagle (Joel Kenny) wrote…

Adding a --purge flag here will free up the "cockroachdb" name in Helm in case a user wants to launch another CRDB cluster using Helm into the same Kubernetes cluster.

Done.

[cockroach-teamcity]

Online preview: http://cockroach-docs-review.s3-website-us-east-1.amazonaws.com/98e1b1672ae873030bd2ab186628a61bb883cef9/

Edited pages:

• v19.1/orchestrate-cockroachdb-with-kubernetes-insecure.md
• v19.2/orchestrate-cockroachdb-with-kubernetes-insecure.md

[DuskEagle]

Reviewed 4 of 6 files at r1, 4 of 4 files at r2.
Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @DuskEagle)