Ensure kind cluster has RFC1123 compliant name

[ghost]

What does this PR do?

This patch:

• Modifies the cluster name generation function to produce - where it previously produced _
• Adds run-time validation that the generated name is compliant
• Renames the function from kubernetesPodName to kubernetesClusterName, reflecting its purpose

Why is it important?

When creating a kind cluster, we must use a cluster name that is a valid Kubernetes resource name, and by extension, a valid DNS name. If not, kind cluster provisioning can fail when kind tries to create resources with invalid names.

For example, if trying to create a cluster called "a_b" (underscores are not permitted), control-plane provisioning will fail with:

host 'a_b-control-plane' must be [...] a valid RFC-1123 DNS subdomain

With this patch, the kubernetes Metricbeat module integration tests are able to run to completion on my system. Without it, kind fails in this manner:

$ kind create cluster --name metricbeat-8_0_0-a34234bc-master
Creating cluster "metricbeat-8_0_0-a34234bc-master" ...
 ✓ Ensuring node image (kindest/node:v1.18.2) 🖼 
 ✓ Preparing nodes 📦  
 ✓ Writing configuration 📜 
 ✗ Starting control-plane 🕹️ 
ERROR: failed to create cluster: failed to init node with kubeadm: command "docker exec --privileged metricbeat-8_0_0-a34234bc-master-control-plane kubeadm init --ignore-preflight-errors=all --config=/kind/kubeadm.conf --skip-token-print --v=6" failed with error: exit status 1
Command Output: I0817 07:52:20.014341     163 initconfiguration.go:200] loading configuration from "/kind/kubeadm.conf"
[config] WARNING: Ignored YAML document with GroupVersionKind kubeadm.k8s.io/v1beta2, Kind=JoinConfiguration
I0817 07:52:20.018540     163 interface.go:400] Looking for default routes with IPv4 addresses
I0817 07:52:20.018553     163 interface.go:405] Default route transits interface "eth0"
I0817 07:52:20.018623     163 interface.go:208] Interface eth0 is up
I0817 07:52:20.018673     163 interface.go:256] Interface "eth0" has 3 addresses :[172.20.0.2/16 fc00:f853:ccd:e793::2/64 fe80::42:acff:fe14:2/64].
I0817 07:52:20.018689     163 interface.go:223] Checking addr  172.20.0.2/16.
I0817 07:52:20.018696     163 interface.go:230] IP found 172.20.0.2
I0817 07:52:20.018702     163 interface.go:262] Found valid IPv4 address 172.20.0.2 for interface "eth0".
I0817 07:52:20.018706     163 interface.go:411] Found active IP 172.20.0.2 
hostport metricbeat-8_0_0-a34234bc-master-control-plane:6443: host 'metricbeat-8_0_0-a34234bc-master-control-plane' must be a valid IP address or a valid RFC-1123 DNS subdomain

Checklist

• My code follows the style guidelines of this project
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have made corresponding change to the default configuration files
• I have added tests that prove my fix is effective or that my feature works
• I have added an entry in CHANGELOG.next.asciidoc or CHANGELOG-developer.next.asciidoc.

How to test this PR locally

Run the integ tests for the kubernetes Metricbeat module with Mage in verbose mode. kind will be seen to run correctly. Without the patch kind will fail to provision the control-plane.

MODULE=kubernetes mage -v integtest

With Mage not in verbose mode, the kubernetes integ test is not run, and no output is seen (because it dies at setup time).

[elasticmachine]

Pinging @elastic/integrations-platforms (Team:Platforms)

[elasticmachine]

💚 Build Succeeded

Expand to view the summary

Build stats

• Build Cause: [Pull request #20627 opened]

• Start Time: 2020-08-17T08:00:35.841+0000

• Duration: 73 min 2 sec

Test stats 🧪

Test	Results
Failed	0
Passed	18002
Skipped	1827
Total	19829

[ghost]

I originally requested @blakerouse for review, but it looks like Blake is taking a quick break.

Would you mind taking a look, @ChrsMark? Thanks.

Note: The Travis test failure is for the test tests/system/test_unix.py::Test::test_unix_with_newline_delimiter on OSX. It's failing with OSError: [Errno 24] Too many open files, which seems to be happening on several different branches.

[ghost]

Looks like @ChrsMark is now on vacation, but @blakerouse is back! :D

[blakerouse]

Change looks good to me, thanks for the fix!

[ghost]

Thanks! I'm not sure about your team's merge/backport procedure. Could you help? Cheers.

[ghost]

What are the next steps for getting this merged? Thanks.

[blakerouse]

@Jarpy Sorry, forgot about merging this. Merged it and created the backport PR for 7.x.

[ghost]

No worries. Many thanks!