Install Obliq AI SRE Agent

Obliq is an AI-Powered infrastructure management platform designed to bring intelligence, automation, and reliability to operations across Kubernetes and AWS environments.

Install Obliq AI SRE Agent using Helm charts.

info

Before you begin installation, ensure that these prerequisites are met.

Install the Agent

This section describes the step-by-step procedure for installing the Obliq AI SRE Agent.

Add the Helm Repository

1. Add the Obliq Charts Helm repository using the following command:

   helm repo add obliq-charts https://repo.obliq.avesha.io/
   helm repo update

   info

   • You can explore the values.yaml Helm chart in Explore the Helm Chart.
   • To learn about parameters of values.yaml, see Helm Chart Parameters.

Create a Registry Secret

Use the following commands to create a registry secret:

# Get credentials from support@aveshasystems.com
kubectl create namespace avesha --dry-run=client -o yaml | kubectl apply -f -
kubectl create secret docker-registry registry \
  --docker-username=YOUR_USERNAME_FROM_AVESHA \
  --docker-password=YOUR_PASSWORD_FROM_AVESHA \
  --docker-email=your-email@company.com \
  --docker-server=https://avesha.azurecr.io/ \
  -n avesha

For advanced configuration, see Secret Management.

Choose an Installation Option

info

To manage environment variables, see .env file setup.

We have provided the following three installation options for you to choose:

• Minimal or Core Services Only
• AWS Integration
• Full Integration

note

If you need to create a custom kubeconfig file, you can download it from your Kubernetes cluster or cloud provider.

Minimal or Core Services Only

Use this installation option for a demonstration or development.

info

For more information, see Kubernetes Permissions.

# Set your OpenAI API key (get from https://platform.openai.com/api-keys)
export OPENAI_API_KEY="sk-your-openai-api-key"

# Install with LoadBalancer for external UI access
helm install obliq-sre-agent obliq-charts/obliq-sre-agent \
  --namespace avesha \
  --create-namespace \
  --set-file global.kubeconfig.content=./kubeconfig `# Path to your kubeconfig file` \
  --set global.env.openai.OPENAI_API_KEY="${OPENAI_API_KEY}" `# Required for AI services` \
  --set avesha-unified-ui.service.type=LoadBalancer `# Expose UI externally` \
  --timeout 15m

AWS Integration

Use this installation option for AWS environments. Before running AWS integration, ensure you have:

• AWS IAM Policies: required IAM roles and policies.
• AWS prerequisites: AWS account configuration and permissions
• Valid AWS credentials: IAM user with required permissions

# Core credentials
export OPENAI_API_KEY="sk-your-openai-api-key"  # Get from OpenAI platform
export AWS_ACCESS_KEY_ID="your-aws-access-key"  # AWS IAM user access key
export AWS_SECRET_ACCESS_KEY="your-aws-secret-key"  # AWS IAM user secret
export AWS_ROLE_ARN_AWS_MCP="arn:aws:iam::123456789012:role/your-aws-mcp-role"  # IAM role for AWS MCP
export AWS_REGION="us-west-2"  # AWS region for resources

# Install with AWS integration and LoadBalancer UI access
helm install obliq-sre-agent obliq-charts/obliq-sre-agent \
  --namespace avesha \
  --create-namespace \
  --set-file global.kubeconfig.content=./kubeconfig `# Path to your kubeconfig` \
  --set global.env.openai.OPENAI_API_KEY="${OPENAI_API_KEY}" `# Required for AI services` \
  --set aws-mcp.enabled=true `# Enable AWS MCP service` \
  --set global.env.aws.AWS_ACCESS_KEY_ID="${AWS_ACCESS_KEY_ID}" `# AWS API access` \
  --set global.env.aws.AWS_SECRET_ACCESS_KEY="${AWS_SECRET_ACCESS_KEY}" `# AWS API secret` \
  --set global.env.aws.AWS_ROLE_ARN_AWS_MCP="${AWS_ROLE_ARN_AWS_MCP}" `# AWS MCP role ARN` \
  --set global.env.aws.AWS_REGION="${AWS_REGION}" `# AWS region` \
  --set avesha-unified-ui.service.type=LoadBalancer `# Expose UI externally` \
  --timeout 15m

Full Integration

Use this installation option for production environments. This option installs all services.

Full integration requires additional setup:

• Kubernetes Permissions: Required RBAC and cluster permissions
• Service Dependencies: Understanding service relationships and dependencies
• Secret Management: Advanced credential and secret handling

For full integration with all services, see the full integration deployment example.

Verify Installation

1. Check if pods are running using the following command:

   kubectl get pods -n avesha

   All pods should be RUNNING.

2. Check the deployment status using the following command:

   # Check deployment status
   helm status obliq-sre-agent -n avesha

3. Verify if core services are ready using the following command:

   kubectl get deployments -n avesha

4. Check for issues (if any) using the following command:

   kubectl get events -n avesha --sort-by='.lastTimestamp' | tail -10

Method to Access the Agent Console

LoadBalancer is the recommended method to access the Agent Console for production environments.

Get the LoadBalancer external IP address using the following command:

# Get the external IP (may take a few minutes to provision)
kubectl get service -n avesha avesha-unified-ui

# Access at: http://<EXTERNAL-IP>:80

After getting the IP address, visit Access the Obliq AI SRE Agent Console. The default login credentials to access the console are:

• Username: admin@aveshasystems.com
• Password: admin123

Alternative Access Methods

Method	Details
NodePort	Add --set avesha-unified-ui.service.type=NodePort instead of LoadBalancer
Port Forward (ClusterIP)	kubectl port-forward -n avesha service/avesha-unified-ui 8080:80

Uninstall the Obliq AI SRE Agent

1. Uninstall the Obliq AI SRE Agent using the following command:

   # Uninstall the application
   helm uninstall obliq-sre-agent -n avesha

2. [Optional] Remove the namespace using the following command:

   # Remove the namespace (optional)
   kubectl delete namespace avesha

3. Remove the cert-manager namespace if you have created it using the following command:

   kubectl delete namespace cert-manager

Troubleshooting

Refer to Obliq AI SRE Agent Installation Issues.