Cloud Pak for AIOps assistant

Cloud Pak for AIOps assistant is available only for technology preview, and is not for production usage.

The Cloud Pak for AIOps assistant is an intelligent assistant that simplifies Cloud Pak for AIOps management by providing on-demand operational guidance, troubleshooting support, and system optimization recommendations.

Prerequisites

1. Enable the feature

Using Red Hat OpenShift user-interface
- Go to the namespace where AIOps is installed.
- Go to Home and select Workloads.
- Click ConfigMaps and select Create ConfigMap.
- Switch to YAML View.
- Paste the following YAML:
```
apiVersion: v1
kind: ConfigMap
metadata:
  name: feature-flag-configmap
data:
  AIOPS_CHAT_ASSISTANT: enabled
```
If a feature-flag-configmap exists, add the AIOPS_CHAT_ASSISTANT flag under data and set it to enabled.
Using command line
- Log in to the platform by using oc login.
- Switch to the project where Cloud Pak for AIOps is installed.
- Create a file named configmap_chat.yaml`, paste the following content, and save it:
```
apiVersion: v1
kind: ConfigMap
metadata:
  name: feature-flag-configmap
data:
  AIOPS_CHAT_ASSISTANT: enabled
```
- Run the command:
```
oc create -f configmap_chat.yaml
```

2. Create the entitlement secret

Get the entitlement key
- Go to the IBM container software library (myibm.ibm.com/products-services/containerlibrary).
- Find your entitlement key and click Copy.
- Save the entitlement key in a safe place so you can use it later when adding the pull secret.
Create the Secret
- You can create the secret by using either the Red Hat OpenShift user-interface or the command line.
- Using Red Hat OpenShift user-interface
  - Log in to the Red Hat OpenShift web-console.
  - Go to Workloads from the left menu and click secrets.
  - Select Create Key/Value Secret.
  - Enter the secret name: aiops-chat-assistant-entitlement-key
  - Enter the key name: entitlementkey
  - Under Value, upload a file that contains the entitlement key.
- Using command line
  - Log in to your Red Hat OpenShift platform by using oc login.
  - Run the following command to switch to the project (namespace) where Cloud Pak for AIOps is installed:
```
oc project <your-namespace>
```
  - Run the following command to create the secret directly with the key value:
```
oc create secret generic aiops-chat-assistant-entitlement-key --from-literal=entitlementkey=<your-key>
```
  - Run the following command to create the secret from a file (example, a text file with the key):
```
oc create secret generic aiops-chat-assistant-entitlement-key --from-file=entitlementkey=/path/to/file
```

3. Configure the Cloud Pak for AIOps chat assistant tile

Chat assistant

Go to AI Model Management and select the Training tab.
Click the Cloud Pak for AIOps Assistant tile.
Click Setup Cloud Pak for AIOps assistant tile.
Under Run model, review the details and click Done.

Enabling chat assistant

After setup is complete and all prerequisites are met, wait a few minutes. The chat widget appears on all Cloud Pak for AIOps user interface pages (except the home page). You can now start interacting with the assistant.

Chat assistant user interface

Sample questions to ask the assistant:
- What are the hardware requirements for Cloud Pak for AIOps production installation?
- How is alert grouping done?
- How to forward AIOps alerts to another tool outside Cloud Pak for AIOps?

Chat assistant response

Add proxy configuration

If there is a proxy configuration setup, the proxy details can be updated in the entitlement secret.

To provide the proxy URL, follow these steps:

Run the following command to encode the proxy URL in Base64:
```
PROXY_URL_BASE64=$(echo -n "<proxy-url-value>" | base64)
```

Run the following command to update the entitlement secret with the proxy URL:

oc patch secret aiops-chat-assistant-entitlement-key -n katamari \
--type='json' \
-p="[{'op': 'add', 'path': '/data/proxyurl', 'value': '$PROXY_URL_BASE64'}]"

Optional: Configure proxy authentication
If the proxy requires authentication, provide the username and password in addition to the proxy URL.
Run the following command to encode the username, password, and enable authentication:
```
PROXY_USERNAME_BASE64=$(echo -n "<proxy-username>" | base64)
PROXY_PASSWORD_BASE64=$(echo -n "<proxy-password>" | base64)
PROXY_ENABLED_BASE64=$(echo -n "enabled" | base64)
```

Update the entitlement secret with authentication values:

oc patch secret aiops-chat-assistant-entitlement-key -n katamari \  --type='json' \  -p="[{'op': 'add', 'path': '/data/proxyusername', 'value': '$PROXY_USERNAME_BASE64'}, \       {'op': 'add', 'path': '/data/proxypassword', 'value': '$PROXY_PASSWORD_BASE64'}, \       {'op': 'add', 'path': '/data/proxyauthenabled', 'value': '$PROXY_ENABLED_BASE64'}]"

Note: Values in Kubernetes secrets must be Base64 encoded. Encode all proxy details before you add them to the entitlement secret. Replace proxy-url-value, proxy-username, and proxy-password with your actual proxy details.

Disable the feature

To disable the feature:

Delete the Cloud Pak for AIOps Chat Assistant training configuration from the AI models user interface.
Set the AIOPS_CHAT_ASSISTANT flag to disable in the feature-flag-configmap.