Added documentation of how client-go interfaces with custom controller code

**What this PR does / why we need it**:

The sample-controller makes extensive use of various mechanisms
available in the client-go library. For writing custom controllers/operators
it will be helpful if there is precise description of how the
client-go library works and how/where it interfaces with
custom controller code.

This patch adds documentation that sheds light on how (parts) of the
client-go library work and its interaction-points with
controller code. The documentation and the diagram comes from [1],
(link included here for reference purpose).

[1] https://medium.com/@cloudark/kubernetes-custom-controllers-b6c7d0668fdf

Contributing the diagram and the writeup was recommended by
by @sttts and @nikhita on https://github.com/kubernetes/sample-controller/issues/13

**Release note**:

```release-note
NONE
```

Kubernetes-commit: 3e22383a8a14eb3bd02630f3f11c0ad7a806e027
This commit is contained in:
Devdatta Kulkarni 2018-04-13 14:32:47 -05:00 committed by Kubernetes Publisher
parent 8a66ce0ffe
commit 220ca987ad
3 changed files with 71 additions and 0 deletions

View File

@ -23,6 +23,13 @@ Changes should not be made to these files manually, and when creating your own
controller based off of this implementation you should not copy these files and
instead run the `update-codegen` script to generate your own.
## Details
The sample controller uses [client-go library](https://github.com/kubernetes/client-go/tree/master/tools/cache) extensively.
The details of interaction points of the sample controller with various mechanisms from this library are
explained [here](docs/controller-client-go.md).
## Purpose
This is an example of how to build a kube-like controller with a single type.

View File

@ -0,0 +1,64 @@
# client-go under the hood
The [client-go](https://github.com/kubernetes/client-go/) library contains various mechanisms that you can use when
developing your custom controllers. These mechanisms are defined in the
[tools/cache folder](https://github.com/kubernetes/client-go/tree/master/tools/cache) of the library.
Here is a pictorial representation showing how the various components in
the client-go library work and their interaction points with the custom
controller code that you will write.
<p align="center">
<img src="images/client-go-controller-interaction.jpeg" height="600" width="700"/>
</p>
## client-go components
* Reflector: A reflector, which is defined in [type *Reflector* inside package *cache*](https://github.com/kubernetes/client-go/blob/master/tools/cache/reflector.go),
watches the Kubernetes API for the specified resource type (kind).
The function in which this is done is *ListAndWatch*.
The watch could be for an in-built resource or it could be for a custom resource.
When the reflector receives notification about existence of new
resource instance through the watch API, it gets the newly created object
using the corresponding listing API and puts it in the Delta Fifo queue
inside the *watchHandler* function.
* Informer: An informer defined in the [base controller inside package *cache*](https://github.com/kubernetes/client-go/blob/master/tools/cache/controller.go) pops objects from the Delta Fifo queue.
The function in which this is done is *processLoop*. The job of this base controller
is to save the object for later retrieval, and to invoke our controller passing it the object.
* Indexer: An indexer provides indexing functionality over objects.
It is defined in [type *Indexer* inside package *cache*](https://github.com/kubernetes/client-go/blob/master/tools/cache/index.go). A typical indexing use-case is to create an index based on object labels. Indexer can
maintain indexes based on several indexing functions.
Indexer uses a thread-safe data store to store objects and their keys.
There is a default function named *MetaNamespaceKeyFunc* defined in [type Store inside package cache](https://github.com/kubernetes/client-go/blob/master/tools/cache/store.go)
that generates an objects key as `<namespace>/<name>` combination for that object.
## Custom Controller components
* Informer reference: This is the reference to the Informer instance that knows
how to work with your custom resource objects. Your custom controller code needs
to create the appropriate Informer.
* Indexer reference: This is the reference to the Indexer instance that knows
how to work with your custom resource objects. Your custom controller code needs
to create this. You will be using this reference for retrieving objects for
later processing.
The base controller in client-go provides the *NewIndexerInformer* function to create Informer and Indexer.
In your code you can either [directly invoke this function](https://github.com/kubernetes/client-go/blob/master/examples/workqueue/main.go#L174) or [use factory methods for creating an informer.](https://github.com/kubernetes/sample-controller/blob/master/main.go#L61)
* Resource Event Handlers: These are the callback functions which will be called by
the Informer when it wants to deliver an object to your controller. The typical
pattern to write these functions is to obtain the dispatched objects key
and enqueue that key in a work queue for further processing.
* Work queue: This is the queue that you create in your controller code to decouple
delivery of an object from its processing. Resource event handler functions are written
to extract the delivered objects key and add that to the work queue.
* Process Item: This is the function that you create in your code which processes items
from the work queue. There can be one or more other functions that do the actual processing.
These functions will typically use the [Indexer reference](https://github.com/kubernetes/client-go/blob/master/examples/workqueue/main.go#L73), or a Listing wrapper to retrieve the object corresponding to the key.

Binary file not shown.

After

Width:  |  Height:  |  Size: 109 KiB