|
|
# HAMi-WebUI
|
|
|
|
|
|
  
|
|
|
|
|
|
## Get Repo Info
|
|
|
|
|
|
```console
|
|
|
helm repo add hami-webui https://project-hami.github.io/HAMi-WebUI
|
|
|
helm repo update
|
|
|
```
|
|
|
|
|
|
_See [helm repo](https://helm.sh/docs/helm/helm_repo/) for command documentation._
|
|
|
|
|
|
## Installing the Chart
|
|
|
|
|
|
Before deploying, ensure that you configure the `values.yaml` file to match your cluster’s requirements. For detailed instructions, refer to the [Configuration Guide for HAMi-WebUI Helm Chart](#configuration-guide-for-hamiwebui-helm-chart)
|
|
|
> _**Important**_: You must adjust the values.yaml before proceeding with the deployment.
|
|
|
|
|
|
Download the `values.yaml` file from the Helm Charts repository:
|
|
|
|
|
|
https://github.com/Project-HAMi/HAMi-WebUI/blob/main/charts/hami-webui/values.yaml
|
|
|
|
|
|
|
|
|
```console
|
|
|
helm install my-hami-webui hami-webui/hami-webui --create-namespace --namespace hami -f values.yaml
|
|
|
```
|
|
|
|
|
|
## Uninstalling the Chart
|
|
|
|
|
|
To uninstall/delete the my-release deployment:
|
|
|
|
|
|
```console
|
|
|
helm delete my-hami-webui
|
|
|
```
|
|
|
|
|
|
The command removes all the Kubernetes components associated with the chart and deletes the release.
|
|
|
|
|
|
## Requirements
|
|
|
|
|
|
| Repository | Name | Version |
|
|
|
|------------|------|---------|
|
|
|
| https://nvidia.github.io/dcgm-exporter/helm-charts | dcgm-exporter | 3.5.0 |
|
|
|
| https://prometheus-community.github.io/helm-charts | kube-prometheus-stack | 62.6.0 |
|
|
|
|
|
|
## Values
|
|
|
|
|
|
| Key | Type | Default | Description |
|
|
|
|-----|------|------------------------------------------------------------------------------------|-------------|
|
|
|
| affinity | object | `{}` | |
|
|
|
| dcgm-exporter.enabled | bool | `true` | |
|
|
|
| dcgm-exporter.nodeSelector.gpu | string | `"on"` | |
|
|
|
| dcgm-exporter.serviceMonitor.additionalLabels.jobRelease | string | `"hami-webui-prometheus"` | |
|
|
|
| dcgm-exporter.serviceMonitor.enabled | bool | `true` | |
|
|
|
| dcgm-exporter.serviceMonitor.honorLabels | bool | `false` | |
|
|
|
| dcgm-exporter.serviceMonitor.interval | string | `"15s"` | |
|
|
|
| externalPrometheus.address | string | `"http://prometheus-kube-prometheus-prometheus.prometheus.svc.cluster.local:9090"` | |
|
|
|
| externalPrometheus.enabled | bool | `false` | |
|
|
|
| fullnameOverride | string | `""` | |
|
|
|
| hamiServiceMonitor.additionalLabels.jobRelease | string | `"hami-webui-prometheus"` | |
|
|
|
| hamiServiceMonitor.enabled | bool | `true` | |
|
|
|
| hamiServiceMonitor.honorLabels | bool | `false` | |
|
|
|
| hamiServiceMonitor.interval | string | `"15s"` | |
|
|
|
| hamiServiceMonitor.relabelings | list | `[]` | |
|
|
|
| hamiServiceMonitor.svcNamespace | string | "kube-system" | Default is "kube-system", but it should be set according to the namespace where the HAMi components are installed. || image.backend.pullPolicy | string | `"IfNotPresent"` | |
|
|
|
| image.backend.repository | string | `"projecthami/hami-webui-be-oss"` | |
|
|
|
| image.backend.tag | string | `"v1.0.0"` | |
|
|
|
| image.frontend.pullPolicy | string | `"IfNotPresent"` | |
|
|
|
| image.frontend.repository | string | `"projecthami/hami-webui-fe-oss"` | |
|
|
|
| image.frontend.tag | string | `"v1.0.0"` | |
|
|
|
| imagePullSecrets | list | `[]` | |
|
|
|
| ingress.annotations | object | `{}` | |
|
|
|
| ingress.className | string | `""` | |
|
|
|
| ingress.enabled | bool | `false` | |
|
|
|
| ingress.hosts[0].host | string | `"chart-example.local"` | |
|
|
|
| ingress.hosts[0].paths[0].path | string | `"/"` | |
|
|
|
| ingress.hosts[0].paths[0].pathType | string | `"ImplementationSpecific"` | |
|
|
|
| ingress.tls | list | `[]` | |
|
|
|
| kube-prometheus-stack.alertmanager.enabled | bool | `false` | |
|
|
|
| kube-prometheus-stack.crds.enabled | bool | `false` | |
|
|
|
| kube-prometheus-stack.defaultRules.create | bool | `false` | |
|
|
|
| kube-prometheus-stack.enabled | bool | `true` | |
|
|
|
| kube-prometheus-stack.grafana.enabled | bool | `false` | |
|
|
|
| kube-prometheus-stack.kubernetesServiceMonitors.enabled | bool | `false` | |
|
|
|
| kube-prometheus-stack.nodeExporter.enabled | bool | `false` | |
|
|
|
| kube-prometheus-stack.prometheus.prometheusSpec.serviceMonitorSelector.matchLabels.jobRelease | string | `"hami-webui-prometheus"` | |
|
|
|
| kube-prometheus-stack.prometheusOperator.enabled | bool | `false` | |
|
|
|
| nameOverride | string | `""` | |
|
|
|
| namespaceOverride | string | `""` | |
|
|
|
| nodeSelector | object | `{}` | |
|
|
|
| podAnnotations | object | `{}` | |
|
|
|
| podSecurityContext | object | `{}` | |
|
|
|
| replicaCount | int | `1` | |
|
|
|
| resources.backend.limits.cpu | string | `"50m"` | |
|
|
|
| resources.backend.limits.memory | string | `"250Mi"` | |
|
|
|
| resources.backend.requests.cpu | string | `"50m"` | |
|
|
|
| resources.backend.requests.memory | string | `"250Mi"` | |
|
|
|
| resources.frontend.limits.cpu | string | `"200m"` | |
|
|
|
| resources.frontend.limits.memory | string | `"500Mi"` | |
|
|
|
| resources.frontend.requests.cpu | string | `"200m"` | |
|
|
|
| resources.frontend.requests.memory | string | `"500Mi"` | |
|
|
|
| securityContext | object | `{}` | |
|
|
|
| service.port | int | `3000` | |
|
|
|
| service.type | string | `"ClusterIP"` | |
|
|
|
| serviceAccount.annotations | object | `{}` | |
|
|
|
| serviceAccount.create | bool | `true` | |
|
|
|
| serviceAccount.name | string | `""` | |
|
|
|
| serviceMonitor.additionalLabels.jobRelease | string | `"hami-webui-prometheus"` | |
|
|
|
| serviceMonitor.enabled | bool | `true` | |
|
|
|
| serviceMonitor.honorLabels | bool | `false` | |
|
|
|
| serviceMonitor.interval | string | `"15s"` | |
|
|
|
| serviceMonitor.relabelings | list | `[]` | |
|
|
|
| tolerations | list | `[]` | |
|
|
|
|
|
|
## Configuration Guide for HAMi-WebUI Helm Chart
|
|
|
|
|
|
### 1. About `dcgm-exporter`
|
|
|
|
|
|
If `dcgm-exporter` is already installed in your cluster, you should disable it by modifying the following setting:
|
|
|
|
|
|
```yaml
|
|
|
dcgm-exporter:
|
|
|
enabled: false
|
|
|
```
|
|
|
This ensures that the existing `dcgm-exporter` instance is used, preventing conflicts.
|
|
|
|
|
|
|
|
|
### 2. About `Prometheus`
|
|
|
|
|
|
#### Scenario 1: If an existing Prometheus is available in your cluster
|
|
|
|
|
|
If your cluster already has a working Prometheus instance, you can enable the external Prometheus configuration and provide the correct address:
|
|
|
|
|
|
```yaml
|
|
|
externalPrometheus:
|
|
|
enabled: true
|
|
|
address: "<your-prometheus-address>"
|
|
|
```
|
|
|
|
|
|
Here, replace <your-prometheus-address> with the actual domain or internal Ingress address for your Prometheus instance.
|
|
|
|
|
|
#### Scenario 2: If no Prometheus or Operator is installed in the cluster
|
|
|
|
|
|
If there is no existing Prometheus or Prometheus Operator in your cluster, you can enable the kube-prometheus-stack to deploy Prometheus:
|
|
|
|
|
|
```yaml
|
|
|
kube-prometheus-stack:
|
|
|
enabled: true
|
|
|
crds:
|
|
|
enabled: true
|
|
|
...
|
|
|
prometheusOperator:
|
|
|
enabled: true
|
|
|
...
|
|
|
```
|
|
|
|
|
|
#### Scenario 3: If Prometheus and Operator already exist, but a separate Prometheus instance is needed
|
|
|
If your cluster has Prometheus and Prometheus Operator, but you want to use a separate instance without affecting the existing setup, modify the configuration as follows:
|
|
|
|
|
|
```yaml
|
|
|
kube-prometheus-stack:
|
|
|
enabled: true
|
|
|
...
|
|
|
```
|
|
|
This allows you to reuse the existing Operator and CRDs while deploying a new Prometheus instance.
|
|
|
|
|
|
### 3. About `jobRelease` Labels
|
|
|
|
|
|
If deploying a completely new Prometheus, you can leave the default `jobRelease: hami-webui-prometheus` unchanged.
|
|
|
|
|
|
***However, if you are integrating with an existing Prometheus instance and modifying the `prometheusSpec.serviceMonitorSelector.matchLabels`, ensure that **all** corresponding `...ServiceMonitor.additionalLabels` are updated to reflect the correct label.***
|
|
|
|
|
|
For example, if you modify:
|
|
|
|
|
|
```yaml
|
|
|
prometheus:
|
|
|
prometheusSpec:
|
|
|
serviceMonitorSelector:
|
|
|
matchLabels:
|
|
|
<jobRelease-label-key>: <jobRelease-label-value>
|
|
|
```
|
|
|
|
|
|
You must also modify all ...ServiceMonitor.additionalLabels in your values.yaml file to match:
|
|
|
|
|
|
```yaml
|
|
|
...ServiceMonitor:
|
|
|
additionalLabels:
|
|
|
<jobRelease-label-key>: <jobRelease-label-value>
|
|
|
```
|
|
|
|
|
|
This ensures that Prometheus will correctly discover all the ServiceMonitor configurations based on the updated labels. |
