feature: install lws controller in llmaz-system namespace

[googs1025]

What this PR does / why we need it

• install lws controller in llmaz-system namespace

Which issue(s) this PR fixes

Fixes #202

Special notes for your reviewer

Does this PR introduce a user-facing change?

 migrate lws controller to llmaz-system namespace.

[googs1025]

helm test:

root@VM-0-12-ubuntu:/home/ubuntu/llmaz# kubectl get pods -nllmaz-system
NAME                                        READY   STATUS    RESTARTS   AGE
llmaz-controller-manager-77978f9964-d8f5z   1/1     Running   0          7m46s
lws-controller-manager-7469fb5d9f-5z5fp     1/1     Running   0          7m46s
lws-controller-manager-7469fb5d9f-6bthh     1/1     Running   0          7m46s
prometheus-llmaz-prometheus-0               2/2     Running   0          7m46s
root@VM-0-12-ubuntu:/home/ubuntu/llmaz# kubectl get pods
NAME                                   READY   STATUS    RESTARTS   AGE
prometheus-operator-55b5c96cf8-q8vhd   1/1     Running   0          4d13h
qwen2-0--5b-0                          1/1     Running   0          5m4s

[googs1025]

This name doesn't seem to be appropriate, please let me know if there is a suitable one. 😄

[kerthcet]

Can we just use a namespace field here?

[kerthcet]

If we want lws to install in lws-system, can we just remove the namespace field from helm charts?

[googs1025]

I want to make it clearer: if we use namespace: "llmaz-system" it will be deployed to llmaz-system, and if we set namespace: "test", then lws-controller will be deployed to the test namespace? In addition: if it is not set, it will be deployed in the lws-system namespace by default? 🤔

like this:

leaderWorkerSet:
  enable: true
  namesapce: "llmaz-system"

[kerthcet]

The simplest solution would be remove the namespace at all, then it will be installed in the same namespace with helm chart, usually the llmaz-system.

Another option is as you typed above, we have a namespace field here.

I'm ok with both approaches.

[kerthcet]

Let's pick the second then, to make it possible with current state -> lws in lws-system namespace.

[googs1025]

for now:

set namespace: "llmaz-system" by default

leaderWorkerSet:
  enable: true
  namespace: "llmaz-system"

and so we can change to other namespace:

leaderWorkerSet:
  enable: true
  namespace: "test-system"

or no set, we will use lws-system

leaderWorkerSet:
  enable: true

[googs1025]

Note: Currently I just use a bool to force installation to llmaz-system or lws-system. This may be a bit hardcoded. If you have other ideas, please let me know.

[googs1025]

/kind feautre

[googs1025]

/kind feature

[kerthcet]

In the future, I would suggest to solve this in the upstream, introduce namespace in helm chart specifically. But good to us now.

[googs1025]

test local:

root@VM-0-12-ubuntu:/home/ubuntu/llmaz# kubectl get pods -A
NAMESPACE            NAME                                             READY   STATUS    RESTARTS   AGE
default              prometheus-operator-55b5c96cf8-q8vhd             1/1     Running   0          7d
default              qwen2-0--5b-0                                    1/1     Running   0          116s
kube-system          coredns-668d6bf9bc-75wmc                         1/1     Running   0          7d
kube-system          coredns-668d6bf9bc-9twhj                         1/1     Running   0          7d
kube-system          etcd-cluster1-control-plane                      1/1     Running   0          7d
kube-system          kindnet-h2gpg                                    1/1     Running   0          7d
kube-system          kindnet-jznjw                                    1/1     Running   0          7d
kube-system          kindnet-sxj8d                                    1/1     Running   0          7d
kube-system          kube-apiserver-cluster1-control-plane            1/1     Running   0          7d
kube-system          kube-controller-manager-cluster1-control-plane   1/1     Running   0          7d
kube-system          kube-proxy-4pqlp                                 1/1     Running   0          7d
kube-system          kube-proxy-8ddvw                                 1/1     Running   0          7d
kube-system          kube-proxy-dv856                                 1/1     Running   0          7d
kube-system          kube-scheduler-cluster1-control-plane            1/1     Running   0          7d
llmaz-system         llmaz-controller-manager-77978f9964-ff7vt        1/1     Running   0          2m16s
llmaz-system         lws-controller-manager-7469fb5d9f-kfv5b          1/1     Running   0          2m16s
llmaz-system         lws-controller-manager-7469fb5d9f-r5mhw          1/1     Running   0          2m16s
llmaz-system         prometheus-llmaz-prometheus-0                    2/2     Running   0          2m15s
local-path-storage   local-path-provisioner-7dc846544d-pgrg7          1/1     Running   0          7d
root@VM-0-12-ubuntu:/home/ubuntu/llmaz# cat chart/values.global.yaml | grep namespace
  namespace: "llmaz-system"

[kerthcet]

could you rebase and and add Action Required: to the user-facing change since we'll need to migrate the lws to another namespace.

[googs1025]

could you rebase and and add Action Required: to the user-facing change since we'll need to migrate the lws to another namespace.

done rebase and release note :)

[kerthcet]

/lgtm
/approve

Thanks!