Support for multi-regional kubernetes clusters

[euroelessar]

Is your feature request related to a problem? Please describe.
We have few k8s clusters spanning across multiple regions, and it is currently impossible to deploy aws-ebs-csi-driver there as::

1. controller supports exactly one region (making it impossible to point it to two regions)
2. controller hard codes driver name (making it impossible to run two independent versions of it)

Describe the solution you'd like in detail
I suggest a combination of the following changes:

1. Node driver can be changed to expose instance's region in accessibility requirements, in addition to currently present AZ there
2. Controller driver can use region from accessibility requirements to use Cloud client for non-default region
3. Controller driver can put region in the generated VolumeID (in a form of $volumeID/$region) so that proper per-region Cloud client can be used in csi requests which have access to VolumeID only (like DeleteVolume)

For backward compatibility, if $region is not present in the VolumeID, we can use "default region" as its value (which is derived using current implementation from AWS_REGION env variable or instance metadata query).

Describe alternatives you've considered

1. Implement own version of csi driver (non-optimal, ideally I'd avoid thinking about this approach)
2. Run different csi drivers per region with dedicated storage classes (complicated ux and blocked by hardcoded driver name)

Additional context

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues and PRs.

This bot triages issues and PRs according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue or PR as fresh with /remove-lifecycle stale
• Mark this issue or PR as rotten with /lifecycle rotten
• Close this issue or PR with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[olemarkus]

Curious what the use case for this is. An instance and its volumes needs to be even in the same AZ. And K8s itself doesn't support multi-region (see e.g kubernetes/cloud-provider#35).

[k8s-triage-robot]

The Kubernetes project currently lacks enough active contributors to adequately respond to all issues and PRs.

This bot triages issues and PRs according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue or PR as fresh with /remove-lifecycle rotten
• Close this issue or PR with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle rotten

[k8s-triage-robot]

The Kubernetes project currently lacks enough active contributors to adequately respond to all issues and PRs.

This bot triages issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Reopen this issue with /reopen
• Mark this issue as fresh with /remove-lifecycle rotten
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/close not-planned

[k8s-ci-robot]

@k8s-triage-robot: Closing this issue, marking it as "Not Planned".

In response to this:

The Kubernetes project currently lacks enough active contributors to adequately respond to all issues and PRs.

This bot triages issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Reopen this issue with /reopen
• Mark this issue as fresh with /remove-lifecycle rotten
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/close not-planned

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.