Deployment of Watson Studio Runtime Usage using the IBM Cloud Pak Deployer
Configure authentication secrets in the vault used by the Cloud Pak Deployer
By default the CP4D monitors codes are pulled from this public Github repository and the CP4D monitors images are pushed into the internal image registry of OpenShift. For this setup no credentials are required to either build the monitor images or push them into the internal OpenShift registry.
Optionally, if the Monitor Github repo is cloned into a private Github registry which does require authentication, a vault secret must be created which is used to authenticate to the private Github repo. For example:
More information can be found here: IBM Cloud Pak Deployer - Monitoring
export STATUS_DIR=<Status directory of the Cloud Pak Deployer>export CONFIG_DIR=<Configuration folder used by the Cloud Pak Deployer># Monitor Source Repo secret./cp-deploy.sh vault set --vault-secret monitors_source_repo_secret --vault-secret-value ThisIsMyGitAccessToken!
Optionally, if the build monitor image should be pushed to an external image registry which does require authentication (so not use the internal OpenShift image registry), a vault secret must be created which is used to authenticate to the external image registry. For example:
export STATUS_DIR=<Status directory of the Cloud Pak Deployer>export CONFIG_DIR=<Configuration folder used by the Cloud Pak Deployer># External Image Registry Credentials./cp-deploy.sh vault set --vault-secret default_monitor_target_cr_user_secret --vault-secret-value TargetCRUserName./cp-deploy.sh vault set --vault-secret monitors_target_cr_password --vault-secret-value TargetCRPassword
Validate the created secrets are available:
export STATUS_DIR=<Status directory of the Cloud Pak Deployer>export CONFIG_DIR=<Configuration folder used by the Cloud Pak Deployer># List all secrets./cp-deploy.sh vault list# Optionally retrieve the values to confirm the correct values are loaded./cp-deploy.sh vault get --vault-secret monitors_source_repo_secret
Add OpenShift Monitoring to the Cloud Pak Deployer configuration
Deploying monitors requires the OpenShift Monitoring to be enabled on the OpenShift cluster. If not already part of the Cloud Pak Deployer configuration, add the openshift_monitoring
configuration. This can be added to an existing yaml configuration file, or a new yaml file can be added to the configuration.
openshift_monitoring:- openshift_cluster_name: <OC_CLUSTER_NAME>user_workload: enabled# Optional properties to persist OpenShift Monitoringpvc_storage_class: ibmc-vpc-block-retain-general-purposepvc_storage_size_gb: 100# Optional Remote Rewrite#remote_rewrite_url: http://remote_rewrite_url:<port>/write# Optional install local instance of Grafana on OpenShift
Property | Description | Mandatory | Allowed values |
---|---|---|---|
openshift_cluster_name | The name of the openshift cluster set | Yes | |
user_workload | Flag to enable or disable OpenShift User workload monitoring | Yes | enabled, disabled |
pvc_storage_class | Storage class used to persist the OpenShift monitoring | No | |
pvc_storage_size_gb | Size of the Persistent Volume Claim for storing OpenShift monitoring | No | |
remote_rewrite_url | Push the Prometheus metrics to a remote system | No | |
grafana_operator | Install the Red Hat Community Grafana Operator | No | enabled, disabled |
grafana_project | Target OpenShift Project for the deployment of Grafana | No |
Add the monitor to the Cloud Pak Deployer configuration
Add the following section to the Cloud Pak Deployer configuration, or add the monitor to an existing monitors
section.
The cp4d_monitors
can be added to an existing configuration yaml file in the config
folder, or create a seperate yaml file:
- OC_PROJECT: The Project name in which Cloud Pak for Data is deployed where the monitors will be added
- OC_CUSTER_NAME: The name of the OpenShift Cluster
# List of cp4d_monitorscp4d_monitors:- name: cp4d-monitor-set-1cp4d_instance: <OC_PROJECT>openshift_cluster_name: <OC_CLUSTER_NAME>default_monitor_source_repo: https://github.com/IBM/cp4d-monitors#default_monitor_source_token_secret: monitors_source_repo_secret#Optional target Container Registry#default_monitor_target_cr: de.icr.io/monitorrepo
Run the Cloud Pak Deployer
Run the Cloud Pak Deployer to implement the changes and deploy the monitors.
export STATUS_DIR=<Status directory of the Cloud Pak Deployer>export CONFIG_DIR=<Configuration folder used by the Cloud Pak Deployer>./cp-deploy.sh env apply