Consider testing on ARM64 machines

[denis-tingaikin]

Currently, NSM supports ARM64 image building, but we're not running tests on ARM64-based machines. To improve scenario coverage, we need to consider adding a new repo or editing an existing one for testing our scenario on ARM64 machines.

Defenition of done

1. Consider the cluster provider with arm64 machines
2. Suggest options of implementation (edit some of integration testing repo or add a new one)
3. Add testing on arm64
4. Make sure that all test work as expected.
5. If some of tests are not working feel free to disable them and open a ticket for each unstable scenario.

Context

networkservicemesh/deployments-k8s#10985
networkservicemesh/deployments-k8s#9215

[glazychev-art]

Options

1. Github macos-14 runner (free)
NOT FEASIBLE AT THIS MOMENT
For the kind-cluster we need to install docker. But we cannot do it, because M1 processor runner doesn't support nested virtualization.
See actions:
https://github.com/douglascamata/setup-docker-macos-action?tab=readme-ov-file#arm64-processors-m1-m2-m3-series-used-on-macos-14-images-are-unsupported
https://github.com/crazy-max/ghaction-setup-docker

Description
https://docs.github.com/en/actions/using-github-hosted-runners/about-github-hosted-runners/about-github-hosted-runners#standard-github-hosted-runners-for-public-repositories
Github has provided a beta version of macos-14 for use (M1).

Virtual Machine 	Processor (CPU) 	Memory (RAM) 	Storage (SSD) 	OS (YAML workflow label)
macOS 	                         3 (M1) 	        7 GB 	   14 GB 	   macos-14 [Beta] 	

Pros: it's free
Cons: possible lack of resources

2. Github macos-xlarge runner
NOT FEASIBLE AT THIS MOMENT. See option 1 above.
https://docs.github.com/en/actions/using-github-hosted-runners/about-larger-runners/about-larger-runners#about-macos-larger-runners

Runner Size	Architecture	Processor (CPU)	       Memory (RAM)	Storage (SSD)	YAML workflow label
XLarge	         arm64 (M1)	6 CPU and 8 GPU	        14 GB	         14 GB	          macos-latest-xlarge, macos-13-xlarge[latest], macos-14-xlarge[Beta]

Pros: more resources
Cons: involves payment
Billing - https://docs.github.com/en/billing/managing-billing-for-github-actions/about-billing-for-github-actions#per-minute-rates

3. Equinix metal
https://deploy.equinix.com/product/servers/c3-large-arm64/
Equinix metal has c3.large.arm64
Pros:

• a lot of resources
• bare metal

Cons: additional configuration may be required when using the cluster API.

4. GKE
GKE provides Tau T2A machine series
We could start with t2a-standard-2
https://cloud.google.com/compute/docs/general-purpose-machines#t2a_machines
Pros: k8s cluster provider
Cons: involves payment

[glazychev-art]

@denis-tingaikin @edwarnicke
I would suggest using Equinix Metal. What do you think?

[denis-tingaikin]

Seems like done.