Add installation configure of LH GuaranteedInstanceManagerCPU

[w13915984028]

Problem:

Per harvester/harvester#5568

Solution:

Update source code and document

Related Issue:
harvester/harvester#5568

Test plan:

Create new cluster, use a harvester-configuration file (step 13 of https://docs.harvesterhci.io/v1.2/install/index) with below items, after installation, check the embedded LH, the related setting is as expected.

https://longhorn.io/docs/1.5.3/references/settings/#guaranteed-instance-manager-cpu

install:
  harvester:
    longhorn:
      default_settings:
        guaranteedInstanceManagerCPU: 6

(valid values are in range [0..12], all other values are converted to 12)

LH chart:

chart/templates/default-setting.yaml:    {{- if not (kindIs "invalid" .Values.defaultSettings.guaranteedInstanceManagerCPU) }}
chart/templates/default-setting.yaml:    guaranteed-instance-manager-cpu: {{ .Values.defaultSettings.guaranteedInstanceManagerCPU }}
chart/README.md:| defaultSettings.guaranteedInstanceManagerCPU | Percentage of the total allocatable CPU resources on each node to be reserved for each instance manager pod when the V1 Data Engine is enabled. The default value is "12". |

** local test **

An expected value 3 of guaranteed-instance-manager-cpu is correctly set to LH via installation configure file.

config file

$ cat harvester-cfg-3.yaml 
scheme_version: 1
install:
  mode: create
  harvester:
    storageClass:
      replicaCount: 1
    longhorn:
      defaultSettings:
        guaranteedReplicaManagerCPU: 4
        guaranteedInstanceManagerCPU: 3

harvester saved config:

$ cat /oem/harvester.config 
schemeversion: 1
serverurl: ""
token: rancher
os:
...
    device: /dev/sda
    configurl: http://192.168.122.159:8000/harvester-cfg-3.yaml
...
    webhooks: []
    addons: {}
    harvester:
        storageclass:
            replicacount: 1
        longhorn:
            defaultsettings:
                guaranteedenginemanagercpu: null
                guaranteedreplicamanagercpu: 4
                guaranteedinstancemanagercpu: 3
        enablegocoverdir: false
    rawdiskimagepath: ""
    persistentpartitionsize: 150Gi

LH settings:

$ kubectl get settings.longhorn.io -A | grep gua

longhorn-system   guaranteed-instance-manager-cpu                                   3                                             35s  // target value

[w13915984028]

@bk201 Kiefer suspect the CI failure may be related to harvester/harvester#5459, will double check with @FrankYang0529

Same failure is found in https://github.com/harvester/harvester-installer/actions/runs/8715101652/job/23906463527

[w13915984028]

@Mergifyio backport v1.2 v1.3

[mergify]

backport v1.2 v1.3

✅ Backports have been created

• #709 Add installation configure of LH GuaranteedInstanceManagerCPU (backport #706) has been created for branch v1.2
• #710 Add installation configure of LH GuaranteedInstanceManagerCPU (backport #706) has been created for branch v1.3

[Vicente-Cheng]

lgtm, thanks!

[w13915984028]

@bk201 The CI passed now.