Skip to content

Commit

Permalink
Write proposal for multi cluster deployment
Browse files Browse the repository at this point in the history
  • Loading branch information
@mcbenjemaa
mcbenjemaa committed Dec 30, 2024
1 parent b5991d9 commit 1baeaa3
Showing 1 changed file with 153 additions and 101 deletions.
254 changes: 153 additions & 101 deletions docs/proposals/multi-proxmox-clusters.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -50,6 +50,7 @@ status: experimental
## Glossary

- **CAPMOX**: Cluster API Provider for Proxmox
- **Proxmox**: Proxmox Virtual Environment

## Summary

Expand All @@ -64,166 +65,217 @@ This proposal aims to enable provisioning Kubernetes clusters on multiple proxmo

## Motivation

This section is for explicitly listing the motivation, goals and non-goals of this proposal.
Proxmox is a great virtualization platform for running Kubernetes clusters, as it provides a simple and easy way to manage virtual machines.
However, it is not possible to provision a single Kubernetes cluster on multiple proxmox clusters.

- Describe why the change is important and the benefits to users.
- The motivation section can optionally provide links to [experience reports](https://github.com/golang/go/wiki/ExperienceReports)
to demonstrate the interest in a proposal within the wider Kubernetes community.
For the following reasons, it is important to enable provisioning Kubernetes clusters on multiple proxmox clusters:

- **Resource Utilization**: By provisioning a single Kubernetes cluster on multiple proxmox clusters, we can better utilize the resources of the proxmox clusters.
- **High Availability**: By provisioning a single Kubernetes cluster on multiple proxmox clusters, we can achieve high availability.

### Goals

- List the specific high-level goals of the proposal.
- How will we know that this has succeeded?
- To design a solution for provisioning Kubernetes Clusters on multiple proxmox clusters.
- To distribute machines to various proxmox clusters.
- To enable high availability of Kubernetes clusters by provisioning them on multiple proxmox clusters.

### Non-Goals/Future Work

- What high-levels are out of scope for this proposal?
- Listing non-goals helps to focus discussion and make progress.
- To manage the lifecycle of the Proxmox clusters instances.
- To manage the security of the Proxmox clusters instances.
- To maintain the state of the Proxmox clusters instances.
- To provide the credentials for the Proxmox clusters instances.

## Proposal

This is where we get down to the nitty gritty of what the proposal actually is.

- What is the plan for implementing this feature?
- What data model changes, additions, or removals are required?
- Provide a scenario, or example.
- Use diagrams to communicate concepts, flows of execution, and states.
This document introduces the concept of a Multi Proxmox Clusters provisioning, a new way to provision HA clusters.
This feature require the user to set a specific credentials and configuration in the Secret and ProxmoxCluster.

[PlantUML](http://plantuml.com) is the preferred tool to generate diagrams,
place your `.plantuml` files under `images/` and run `make diagrams` from the docs folder.

### User Stories

- Detail the things that people will be able to do if this proposal is implemented.
- Include as much detail as possible so that people can understand the "how" of the system.
- The goal here is to make this feel real for users without getting bogged down.

#### Story 1

As a developer or Cluster operator, I would like to Provision K8s Cluster on multiple Proxmox Clusters,
so that I can better utilize the resources of the proxmox clusters.

#### Story 2

### Requirements (Optional)
As a developer or Cluster operator, I would like CAPMOX to expose configurations to enable provisioning Kubernetes clusters on multiple proxmox clusters.

#### Story 3

As a developer or Cluster operator, I would like CAPMOX to use a specific schema for Proxmox credentials that is used to provision the Kubernetes cluster on multiple proxmox clusters.

#### Implementation Details/Notes/Constraints

The proposed solution is designed to make the provision of workload clusters as easy as possible while keeping backward compatible with the existing CAPMOX.

Some authors may wish to use requirements in addition to user stories.
Technical requirements should derived from user stories, and provide a trace from
use case to design, implementation and test case. Requirements can be prioritised
using the MoSCoW (MUST, SHOULD, COULD, WON'T) criteria.
##### ProxmoxCluster CRD

The FR and NFR notation is intended to be used as cross-references across a CAEP.
We propose to introduce a new field `spec.settings` in ProxmoxCluster CRD.
The goal of this field is to determine which mode to use for provisioning the Kubernetes cluster.


```yaml
apiVersion: infrastructure.cluster.x-k8s.io/v1alpha1
kind: ProxmoxCluster
metadata:
name: test
spec:
settings:
mode: Default
instances:
- name: pmox-txl
nodes: [pmox-txl-1, pmox-txl-2]
template:
sourceNode: pmox-txl-1
templateID: 1000
credentialsRef:
name: "pmox-txl-proxmox-credentials"
- name: pmox-fra
nodes: [pmox-fra-1, pmox-fra-2, pmox-fra-3]
template:
templateSelector:
matchTags:
- capi
- v1.30.5
credentialsRef:
name: "pmox-fra-proxmox-credentials"
controlPlaneEndpoint:
host: 10.0.0.10
port: 6443
ipv4Config:
addresses: [10.0.0.11-10.0.0.50]
prefix: 24
gateway: 10.0.0.1
dnsServers: [10.0.0.1]
allowedNodes: []
credentialsRef:
name: "test-proxmox-credentials"
```
The difference between goals and requirements is that between an executive summary
and the body of a document. Each requirement should be in support of a goal,
but narrowly scoped in a way that is verifiable or ideally - testable.
The `spec.settings.mode` field will have the following values:
* `Default`: This is the default mode. It will provision the Kubernetes cluster on a single proxmox cluster.
* `SingleInstance`: This mode will provision the Kubernetes cluster on a single proxmox cluster.
* `MultiInstance`: This mode will provision the Kubernetes cluster on multiple proxmox clusters.

#### Functional Requirements
The `spec.settings.instances` field will have the list of proxmox clusters to provision the Kubernetes cluster on.

Functional requirements are the properties that this design should include.
* `spec.settings.mode=Default`: if this mode is set and the `spec.settings.instances` field is not set, the Kubernetes cluster will be provisioned on the default proxmox cluster,
and when instances is set it will provision the Kubernetes cluster on the first proxmox instance.

##### FR1
* `spec.settings.mode=SingleInstance`: if this mode is set, the Kubernetes cluster will be provisioned on a single proxmox cluster chosen randomly from the list of proxmox clusters in the `spec.settings.instances` field.

##### FR2
* `spec.settings.mode=MultiInstance`: if this mode is set, the Kubernetes cluster will be provisioned on multiple proxmox clusters from the list of proxmox clusters in the `spec.settings.instances` field.

#### Non-Functional Requirements
Notes:

Non-functional requirements are user expectations of the solution. Include
considerations for performance, reliability and security.
* `spec.allowedNodes` can be deprecated and have no effect.
* `spec.credentialsRef` will be used only for the default proxmox cluster (can be deprecated in future).
* If no `mode` is set the Cluster will be provisioned on the default proxmox cluster.

##### NFR1
This will make sure to distribute the machines to the different proxmox clusters.

##### NFR2
##### ProxmoxMachine CRD

### Implementation Details/Notes/Constraints
First, in order to enable multi cluster deployment, we need the feature of selecting the template by tags https://github.com/ionos-cloud/cluster-api-provider-proxmox/pull/343
The tags, should match tags set on VM templates in Proxmox for all instances.

- What are some important details that didn't come across above.
- What are the caveats to the implementation?
- Go in to as much detail as necessary here.
- Talk about core concepts and how they relate.
In order to have a way of manually setting an instance in machine level for example, when using a MachineDeployment.
we can add the field `instance` to the ProxmoxMachine CRD.

### Security Model
```yaml
kind: ProxmoxMachineTemplate
apiVersion: infrastructure.cluster.x-k8s.io/v1alpha1
metadata:
name: "test-worker-1"
spec:
template:
spec:
templateSelector:
matchTags:
- capi
- v1.30.5
instance: pmox-txl
```

Document the intended security model for the proposal, including implications
on the Kubernetes RBAC model. Questions you may want to answer include:
##### Proxmox Client

* Does this proposal implement security controls or require the need to do so?
* If so, consider describing the different roles and permissions with tables.
* Are their adequate security warnings where appropriate (see https://adam.shostack.org/ReederEtAl_NEATatMicrosoft.pdf for guidance).
* Are regex expressions going to be used, and are their appropriate defenses against DOS.
* Is any sensitive data being stored in a secret, and only exists for as long as necessary?
The new model of having multiple proxmox clusters will require changes in the Proxmox client to support multiple proxmox clusters.

### Risks and Mitigations
###### The client interface

- What are the risks of this proposal and how do we mitigate? Think broadly.
- How will UX be reviewed and by whom?
- How will security be reviewed and by whom?
- Consider including folks that also work outside the SIG or subproject.
The client interface will stay the same, any required changes will be adapted.

## Alternatives
###### The client implementation

The `Alternatives` section is used to highlight and record other possible approaches to delivering the value proposed by a proposal.
The api client struct will now have a map of clients instead of a single client.
The map will have the name of the proxmox cluster as the key and the client as the value.

## Upgrade Strategy
```go
// APIClient Proxmox API client object.
type APIClient struct {
clients map[string]*proxmox.Client
logger logr.Logger
}
```

* In default mode, the client will use the default client key.

The function `NewAPIClient()` will be initialized in the controller.

If applicable, how will the component be upgraded? Make sure this is in the test plan.
The controller will determine which instance to use and provide the name to the APIClient.
The APIClient functions should have a param that specify which instance to use, based on instance name.

Consider the following in developing an upgrade strategy for this enhancement:
- What changes (in invocations, configurations, API use, etc.) is an existing cluster required to make on upgrade in order to keep previous behavior?
- What changes (in invocations, configurations, API use, etc.) is an existing cluster required to make on upgrade in order to make use of the enhancement?
```go
## Additional Details
// selectNextInstance selects a client based on the request.
func (c *APIClient) CloneVM(, templateID int, clone capmox.VMCloneRequest, instance string) (capmox.VMCloneResponse, error) {
}
```

### Test Plan [optional]

**Note:** *Section not required until targeted at a release.*
**State**
The proxmox instance will be saved in ProxmoxMachine as `spec.instance`.
The state of distributing the machines will be saved in the ProxmoxCluster as `status.instances`.

Consider the following in developing a test plan for this enhancement:
- Will there be e2e and integration tests, in addition to unit tests?
- How will it be tested in isolation vs with other components?

No need to outline all of the test cases, just the general strategy.
Anything that would count as tricky in the implementation and anything particularly challenging to test should be called out.
### Security Model

All code is expected to have adequate tests (eventually with coverage expectations).
Please adhere to the [Kubernetes testing guidelines][testing-guidelines] when drafting this test plan.
This proposal will add a new fields to ProxmoxCluster and ProxmoxMachine CRDs.
The permissions will stay the same.
This feature will also may require the user to manage multiple secrets.

[testing-guidelines]: https://git.k8s.io/community/contributors/devel/sig-testing/testing.md
### Risks and Mitigations

### Graduation Criteria [optional]
* The new feature introduce new secrets for proxmox credentials.

**Note:** *Section not required until targeted at a release.*
## Alternatives

Define graduation milestones.
The alternative is to use the existing CAPMOX and provision the Kubernetes cluster on a single proxmox cluster.

These may be defined in terms of API maturity, or as something else. Initial proposal should keep
this high-level with a focus on what signals will be looked at to determine graduation.
## Upgrade Strategy

Consider the following in developing the graduation criteria for this enhancement:
- [Maturity levels (`alpha`, `beta`, `stable`)][maturity-levels]
- [Deprecation policy][deprecation-policy]
since there are changes within the API and/or deprecations, we need to create a new version.
`v1alpha2` or `v1alpha3` will be the new version of the ProxmoxCluster and ProxmoxMachine CRDs.

Clearly define what graduation means by either linking to the [API doc definition](https://kubernetes.io/docs/concepts/overview/kubernetes-api/#api-versioning),
or by redefining what graduation means.
CAPMOX will support multiple versions at the same time.
only when the end of life of the the new version must be communicated with the users.

In general, we try to use the same stages (alpha, beta, GA), regardless how the functionality is accessed.
## Additional Details

[maturity-levels]: https://git.k8s.io/community/contributors/devel/sig-architecture/api_changes.md#alpha-beta-and-stable-versions
[deprecation-policy]: https://kubernetes.io/docs/reference/using-api/deprecation-policy/
### Test Plan

### Version Skew Strategy [optional]
- Unit test coverage for the Proxmox instance selection and distributions.
- E2e tests if possible, to test the distribution of the machines to the different proxmox clusters.

If applicable, how will the component handle version skew with other components? What are the guarantees? Make sure
this is in the test plan.
### Graduation Criteria

Consider the following in developing a version skew strategy for this enhancement:
- Does this enhancement involve coordinating behavior in the control plane and in the kubelet? How does an n-2 kubelet without this feature available behave when this feature is used?
- Will any other components on the node change? For example, changes to CSI, CRI or CNI may require updating that component before the kubelet.
This feature will be released in a new version CRD `v1alpha3` to CAPMOX.

## Implementation History

- [ ] MM/DD/YYYY: Proposed idea in an issue or [community meeting]
- [ ] MM/DD/YYYY: Compile a Google Doc following the CAEP template (link here)
- [ ] MM/DD/YYYY: First round of feedback from community
- [ ] MM/DD/YYYY: Present proposal at a [community meeting]
- [ ] MM/DD/YYYY: Open proposal PR

<!-- Links -->
[community meeting]: https://docs.google.com/document/d/1ushaVqAKYnZ2VN_aa3GyKlS4kEd6bSug13xaXOakAQI/edit#heading=h.pxsq37pzkbdq
- [ ] 2024-12-19: Open proposal PR
- [ ] MM/DD/YYYY: Present proposal at a community meeting

0 comments on commit 1baeaa3

Please sign in to comment.