An example service-configuration.yaml file is available here. The yaml file includes the following fields.
There are only 2 mandatory configuration for admin to configure in service-configuration.yaml. And see the yaml format data following.
rest-server:
# database admin username
default-pai-admin-username: your_default_pai_admin_username
# database admin password
default-pai-admin-password: your_default_pai_admin_passwordBesides the default cofiguration to make OpenPAI start, admin could customize each service component within permissible scope. From the example serivce-configuration.yaml, you could find a lot of commented fileds, such as the following. For example, If you wanna customize drivers, you can uncomment the filed, and overwrite the default value with your expected value.
#drivers:
# set-nvidia-runtime: false
# # You can set drivers version here. If this value is miss, default value will be 384.111
# # Current supported version list
# # 384.111
# # 390.25
# # 410.73
# version: "384.111"According to your requirements, choose the component which you wanna customized. Please read the table and the link in the table carefully.
| Service | Description | Tutorial |
|---|---|---|
| cluster | Configure data-path, cluster-id, azure rdma switch and docker-registry to pull image. | Link |
| drivers | Configure drivers version and nvidia runtime. | Link |
| hadoop-resource-manager | yarn exporter port and default vc configuration | Link |
| yarn-frameworklauncher | frameworklauncher port configuration | Link |
| rest-server | admin account, github for marketplace, jwt-secret | Link |
| webportal | webportal port configuration | Link |
| grafana | grafana port configuration | Link |
| node-exporter | node-exporter port configuration | Link |
| alert-manager | port configuration and alerts email configuration | Link |
| prometheus | port configuration and scrape interval configuration | Link |
| pylon | port configuration | Link |