-
Notifications
You must be signed in to change notification settings - Fork 8
(architecture): add Fabric Management page #71
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: master
Are you sure you want to change the base?
Conversation
🚀 Deployed on https://preview-71--hedgehog-docs.netlify.app |
d74cb91
to
92c9af4
Compare
92c9af4
to
c5b059c
Compare
c5b059c
to
25c0ab0
Compare
this looks good @pau-hedgehog, my only concern is that I feel there's a big overlap between the first section in this PR and what you previously added in the architecture overview section, and I wonder whether we should just collapse the two sections into one. That being said, having more documentation is never a bad thing, so I'm not against adding this if you / other people feel that they're better kept separate. |
I support adding it. I can see that a user might say, what happens when I do X, Y or Z. They would have those questions answered in the Fabric Management Workflow. |
3aa7d55
to
84a529a
Compare
84a529a
to
6c99034
Compare
I've made some amendments to try to make both @edipascale and @mrbojangles3 happy ;) |
docs/architecture/management.md
Outdated
|
||
### **kubectl & kubectl fabric** | ||
- [`kubectl`](https://kubernetes.io/docs/reference/kubectl/) is the standard CLI tool for [Kubernetes](https://kubernetes.io). | ||
- `kubectl fabric` is a plugin that extends `kubectl` with fabric-specific commands and interacts with the Kubernetes API Server. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
As you're going into that details in k9s section it would be good to describe fabric plugin capability to do inspects, ssh/reboot/reinstall here as well to avoid creating an impression that it's only available through the k9s
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm adding:
It supports the following commands:
- vpc – Manage VPCs: create, attach, peer, or wipe all related resources
- switch – Manage switches: get IP, SSH, serial, reboot, reinstall
- connection – View connections: management, fabric, vpc-loopback
- switchgroup – Create SwitchGroups
- external – Manage external BGP peers and VPC peering
- wiring – Export fabric wiring diagrams
- inspect – Inspect and troubleshoot fabric resources:
- fabric
– Overview of control nodes and switches
- switch
, port
– Status, counters, and usage
- server
, connection
– VPC/external attachments and link details
- vpc
– Subnet attachments and reachability
- bgp
, lldp
– Neighbor status checks
- ip
, mac
– Lookup usage across the fabric
- access
– Check reachability between VPCs, servers, and IPs
docs/architecture/management.md
Outdated
|
||
--- | ||
|
||
## **GitOps Functionality (ArgoCD)** |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It's a bit misleading this way. Please explain GitOps in general and that it could be used with any tools that supports K8s (e.g. ArgoCD, Flux, etc.)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Good point. Changing to:
GitOps Integration
GitOps is a continuous deployment approach where the desired state of a Kubernetes cluster is stored in Git. A GitOps controller monitors this repository and ensures the actual cluster state matches the declared state, using the Kubernetes API.
Hedgehog Fabric is fully compatible with GitOps workflows and can be managed declaratively using any GitOps-compatible tool such as ArgoCD or Flux.
-
User Responsibilities:
- Define Fabric Custom Resources (CRs) in YAML.
- Commit and push these CRs to a Git repository for version control and collaboration.
-
GitOps Tool Responsibilities:
- Monitor the Git repository for updates.
- Apply changes to the cluster automatically via the Kubernetes API Server.
This enables seamless integration of Hedgehog Fabric into modern CI/CD pipelines and Git-centric workflows.
docs/architecture/management.md
Outdated
|
||
GitOps workflows can be leveraged using [ArgoCD](https://argo-cd.readthedocs.io/en/stable/). This is an alternative approach to show that a Fabric can be used with industry standard tools seamlessly. | ||
|
||
- **User Actions:** |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Could go as an example workflow with ArgoCD
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
refactored the GitOps section. See comment above
|
||
The installation of a Hedgehog Fabric deployment is carried out using Fabricator (hhfab CLI). Once deployed, ongoing operations are managed via the [Kubernetes](https://kubernetes.io) CLI, [`kubectl`](https://kubernetes.io/docs/reference/kubectl/). | ||
|
||
In this workflow, the Kubernetes API Server processes the Fabric Custom Resources (CRs) and forwards them to the Fabric Controller. The Fabric Controller then translates these high-level network intents into concrete network configurations and manages their application to the underlying infrastructure. The Controller continuously monitors the state of the network resources, ensuring they match the desired state defined in the CRs. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
API Server isn't forwarding CRs, controllers watches API for CR changes
docs/architecture/management.md
Outdated
Fabricator -->|Applies CRs| K8S | ||
KubectlFabric -->|Interacts with| K8S | ||
K9s -->|UI Manages| K8S | ||
K8S -->|Sends CRDs| FC |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
API Server isn't forwarding CRs, controllers watches API for CR changes
docs/architecture/management.md
Outdated
User -.->|CLI| Fabricator | ||
User -.->|CLI| Kubectl | ||
Kubectl -.->|Plugin| KubectlFabric | ||
Fabricator -->|SSH access| K9s |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not sure what is represented by this, could you please elaborate?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is how you usually access k9s, but maybe it's too detailed for what we want to display
hhfab vlab ssh -n control-1
k9s
docs/architecture/management.md
Outdated
|
||
### **User** | ||
- **Creates Fabric CR YAMLs** and applies them through standard Kubernetes resource management. | ||
- **Directly interacts with SONiC switches** via the Fabricator CLI. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I guess it's about hhfab vlab ssh
helper so it should be removed from here. Additionally, we don't allow users to do the changes on the switches
docs/architecture/management.md
Outdated
- `kubectl fabric` is a plugin that extends `kubectl` with fabric-specific commands and interacts with the Kubernetes API Server. | ||
|
||
### **Fabricator** | ||
- CLI tool that provides direct interaction with the Kubernetes API. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It doesn't provide it, what do you mean?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
At that time I was thinking of hhfab init -w
but it's not the expected workflow. I'm removing it
docs/architecture/management.md
Outdated
|
||
### **Fabricator** | ||
- CLI tool that provides direct interaction with the Kubernetes API. | ||
- Can apply configurations via **SSH access** (using **K9s**) or by directly managing Fabric CRs with YAML files. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Kinda same - I guess you're talking about hhfab vlab ssh
which shouldn't be mentioned outside of the vlab section of the docs
Signed-off-by: Pau Capdevila <[email protected]>
Signed-off-by: Pau Capdevila <[email protected]>
Signed-off-by: Pau Capdevila <[email protected]>
6c99034
to
88b640e
Compare
Signed-off-by: Pau Capdevila <[email protected]>
Signed-off-by: Pau Capdevila <[email protected]>
44fe910
to
77b16ea
Compare
77b16ea
to
103bfe4
Compare
Signed-off-by: Pau Capdevila <[email protected]>
103bfe4
to
857ac1d
Compare
|
||
--- | ||
|
||
## **Management Workflow Overview** |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I didn't go through the whole page, this is just a note in passing - I'd strongly recommend against manually formatting all section/subsection headings in bold. The formatting from the theme should be enough already, no need to add manual formatting; if we decided it's not enough, this is something we should change by tuning the CSS for the whole website, for consistency.
No description provided.