Robusta¶
Optional - Requires Robusta SaaS
This toolset is NOT enabled by default. It requires integration with the Robusta SaaS platform and proper authentication credentials.
The Robusta toolset provides advanced observability capabilities by connecting HolmesGPT to the Robusta SaaS platform. When enabled, it gives HolmesGPT access to historical data, change tracking, and resource recommendations that are not available from standard Kubernetes APIs.
Prerequisites¶
To use this toolset, you need:
- An active Robusta SaaS account
- Valid authentication credentials (provided via environment variables or configuration file)
- The Robusta platform deployed in your cluster
What It Adds¶
When connected to Robusta SaaS, HolmesGPT gains access to:
- Historical Alert Data: Fetch detailed metadata about past alerts and incidents, including context that may no longer be available in Prometheus or AlertManager
- Change Tracking: Query configuration changes across your entire cluster within specific time ranges, helping identify what changed before an incident
- Resource Recommendations: Get AI-powered recommendations for resource requests and limits based on actual historical usage patterns
Configuration¶
The toolset requires authentication to Robusta SaaS. You can provide credentials in three ways:
Option 1: Automatic (via Robusta Helm Chart)¶
If you deploy HolmesGPT as part of the Robusta Helm chart, credentials are automatically configured. The Helm chart handles mounting the necessary secrets and configuration files.
Option 2: Environment Variables¶
export ROBUSTA_UI_TOKEN="<base64-encoded-token>"
# OR provide individual credentials:
export ROBUSTA_ACCOUNT_ID="<account-id>"
export STORE_URL="<store-url>"
export STORE_API_KEY="<api-key>"
export STORE_EMAIL="<email>"
export STORE_PASSWORD="<password>"
Option 3: Configuration File¶
The toolset will automatically look for credentials in /etc/robusta/config/active_playbooks.yaml
(or the path specified by ROBUSTA_CONFIG_PATH
).
Enabling the Toolset¶
Capabilities¶
Tool Name | Description |
---|---|
fetch_finding_by_id | Fetches detailed metadata about a specific Robusta finding (alerts, deployment updates, etc.) including historical context |
fetch_configuration_changes | Retrieves all configuration changes in a given time range, optionally filtered by namespace or workload |
fetch_resource_recommendation | Provides resource optimization recommendations based on actual historical usage for Deployments, StatefulSets, DaemonSets, and Jobs |
Use Cases¶
This toolset is particularly useful for:
- Root Cause Analysis: Understanding what configuration changes occurred before an incident
- Resource Optimization: Getting data-driven recommendations for right-sizing workloads
- Historical Investigation: Accessing alert context and metadata that may have been lost or expired in other systems
- Change Management: Tracking who changed what and when across your infrastructure
Notes¶
- The toolset will only be functional if valid Robusta SaaS credentials are provided
- If credentials are missing or invalid, the toolset will be disabled automatically
- This integration provides read-only access to your Robusta data