Power down & de-provision old equinix metal arm64 ci runners

[jmhbnz]

What would you like to be added?

In #16801 we moved to managed arm64 runners which has allowed us to run more arm64 tests for every pull request and has all round been a great improvement.

We have some resource tuning still to do but while these managed runners are working for us I suggest we power down the two old arm64 boxes which are:

Hostname	IP	Operating System	Region
etcd-c3-large-arm64-runner-01	86.109.7.233	Ubuntu 22.04.1 LTS	Washington DC
etcd-c3-large-arm64-runner-02	147.28.151.226	Ubuntu 22.04.1 LTS	Washington DC

Why is this needed?

Reduce unnecessary compute resource consumption. We can always power them back on if we need to start using them again.

[jmhbnz]

Will leave this for a few days in case anyone has a view on it before powering off the machines via the equinix console.

[jmhbnz]

Machines have been powered off.

Note: We will still want to de-provision these two machines at some point in the future as powering them off does not stop billing for the machines. Given that will leave this issue open to track the de-provisioning in future.

[jmhbnz]

@etcd-io/maintainers-etcd I plan on de-provisioning these old servers on 15th January. Please respond here if you have any objection to this.

We have been operating arm64 ci via actuated managed runners for a while now successfully. Additionally should the need arise we have the required access to create new machines ourselves via the equinix dashboard so I don't see any problem as it is trivial to spin a new machine and setup github actions runner if we ever need to in future.

This would reduce some wasted spend of cncf equinix credits.

[jmhbnz]

Closing - Both old servers have been deleted via the equinix metal web console.

In future we retain access to the equinix web console and recreate new servers on demand if required.