Skip to main content

​
Your First 15 Minutes with RubixKube

Welcome! You’ve installed RubixKube and connected your cluster. Now let’s explore the interface and see what RubixKube can do. This tutorial takes about ** 15 minutes** and gives you a solid foundation.
Prerequisites: - RubixKube account created

​
Step 1: Understanding the Dashboard

After logging in, you’ll land on the Dashboard - your command center for infrastructure reliability.
RubixKube Dashboard overview

​
Main Metrics at a Glance

The dashboard shows 4 key metrics:

System Health

​
100% = Healthy

Overall health of your infrastructure based on pod status, resource usage, and incident history

Active Insights

​
Number of issues detected

Current problems RubixKube has identified that need attention (pods failing, resource issues, etc.)

Intelligent Analysis

​
RCA Reports generated

Root Cause Analysis reports created by the RCA Pipeline Agent

Agents

​
3/3 Active

Status of AI agents monitoring your cluster (Observer, RCA Pipeline, Memory, SRI)

​
Step 2: Check Your Cluster Connection

Scroll down to the Infrastructure panel to verify your cluster is connected:
Connected cluster showing resources
What you’ll see: - ** What you’ll see:** status (green indicator)
  • Node count - How many nodes in your cluster
  • Pod count - Total pods running
  • Deployments - Number of deployments
  • Services - Exposed services
First time? If you just installed, you might see low numbers. Deploy some workloads to see RubixKube monitor them!

​
Step 3: Explore the Navigation

RubixKube’s interface is organized into logical sections:

​
Core Section

What: Overview of system health and recent activity
When to use: Daily check-in, incident triage
Key info: Metrics, insights, activity feed
What: Conversational interface with SRI Agent
When to use: Query your infrastructure using natural language
Try asking: “What pods are failing?” or “Show me memory usage”

​
Monitoring Section

What: Detailed view of all detected issues
When to use: Investigate specific incidents
Shows: Incident details, RCA analysis, suggestions
What: Topology and resource visualization
When to use: Understand service dependencies
Shows: Infrastructure graph, resource relationships

​
Management Section

What: List of connected Kubernetes clusters
When to use: Manage multi-cluster setups
Shows: Cluster health, version, resources
What: Status of all AI agents
When to use: Verify agents are active and healthy
Shows: Agent list, capabilities, last activity
What: Custom dashboards and visualizations
When to use: Create personalized views
Status: Beta feature, coming soon
What: Third-party tool connections
When to use: Configure Slack, Prometheus, etc.
Shows: 17 available integrations
What: Organization and user configuration
When to use: Update profile, preferences, API settings
Tabs: Organization, Preferences, System, Security

​
Step 4: View Active Agents

Navigate to Agents to see your AI team:
Active AI agents in RubixKube

​
You should see 3 active agents:

AgentRoleStatus
RCA Pipeline AgentAnalyzes incidents and generates root cause reportsActive
RubixKube Observer AgentMonitors your cluster continuouslyActive
SRI AgentProvides conversational interface via ChatActive
All agents active = RubixKube is fully operational! They work together to detect, analyze, and help you resolve issues.

​
Step 5: Try the Chat Interface

Click Chat in the navigation to access the conversational interface:
RubixKube Chat interface with SRI Agent

​
Quick Actions Available

The Chat interface offers pre-built action categories:

Infrastructure

Manage clusters, nodes, and resources

Security

Security analysis and compliance checks

Monitoring

Performance metrics and alerts

Troubleshoot

Debug issues and analyze logs

Networking

Network policies and connectivity

​
Try Your First Query

Type a question in the chat box: Example queries to try: ``` “What pods are running in my cluster?” “Show me any failing pods” “What’s the memory usage across all pods?” “Are there any recent events I should know about?” 

<Tip>
<strong>Pro Tip:</strong> Use natural language! The SRI Agent understands questions the way you'd ask a colleague.
</Tip>

---

## Step 6: Understanding the Activity Feed

Back on the **Dashboard** , the **Dashboard** shows recent events:

**What you'll see:** -** What you'll see:** - Pod crashes, resource issues
-**Insights generated**  - AI analysis of problems
-**Resolution events**  - When issues are fixed
-**Configuration changes**  - Deployments, scaling events

**Example Activity:** ```
 Container experiencing repeated crashes in crash-loop-demo
   Severity: medium | Status: active | 2 minutes ago

 Out of memory (OOMKilled) detected on Pod/memory-hog-demo
   Severity: HIGH | Status: active | 2 minutes ago
Activity Feed is live - it updates in real-time as RubixKube detects issues. No refresh needed!

​
Step 7: Check the RCA Live Stream

The RCA Live Stream shows the RCA Pipeline Agent working in real-time: What you’ll see: - Function calls being made
  • Analysis in progress
  • Evidence gathering
  • Pattern matching

​
This is the AI at work!

When RubixKube detects an incident, the RCA Pipeline Agent immediately starts analyzing:1.Gathering evidence - Logs, metrics, events 2.Correlating data - Finding relationships 3.Identifying root cause - Determining what went wrong 4.Generating suggestions - Proposing fixesThis usually takes ** 30-90 seconds** . You’ll see the progress percentage.

​
Step 8: Explore Settings

Navigate to Settings to configure your RubixKube instance:
RubixKube Settings page

​
Available tabs:

  • Organization
  • Preferences
  • System
  • Security
  • Integrations
  • Workspace name and plan
  • Active users and clusters
  • Contact information
  • User profile details
  • Account role and permissions

​
What You’ve Learned

After these first steps, you now know:

Dashboard Navigation

How to navigate RubixKube’s interface and find what you need

Key Metrics

What System Health, Insights, and Agents mean

Chat Interface

How to query your infrastructure using natural language

Agent System

Which AI agents are monitoring your cluster

​
Next Steps

Now that you understand the basics, try these tutorials:

Detect ImagePullBackOff

See how RubixKube detects container image issues

Detect OOMKilled Pods

Watch memory analysis and detection in action

Chat with Your Cluster

Learn advanced Chat queries and tips

Understand Core Concepts

Deep dive into how RubixKube works

​
Common Questions

Answer: If everything is healthy, you won’t see insights! RubixKube only alerts on actual issues.To test: Deploy our tutorial scenarios to see detection in action.
Answer: Real-time via WebSocket connection. You’ll see updates within seconds of changes.Manual refresh: Click the “Refresh all data” button.
Answer: The Workspace feature (coming soon) will let you create custom dashboards.Beta: Use the Chat to query specific data you want to see.
Answer: Total events detected by RubixKube. Click the notification bell to see details.Includes: Incidents, insights, agent activity, system events.

​
Pro Tips

Use Keyboard Shortcuts

  • ⌘K - Open search
  • Quick navigation to any page

Bookmark Important Views

  • Dashboard for daily check-ins
  • Insights for incident review
  • Chat for quick queries

Check Agents Daily

Ensure all 3 agents show “Just now” for last activity

Explore Integrations

Connect Slack for notifications or Prometheus for enhanced metrics

​
Troubleshooting

Possible causes: - Observer agent not connected
  • Network connectivity issue
  • Browser cache problem
Solutions: 1. Check observer pod: kubectl get pods -n rubixkube 2. Verify it’s running (STATUS = Running) 3. Clear browser cache and refresh 4. Try different browser
Check:
  1. Navigate to Agents page
  2. Look for “Last active” timestamp
  3. If >5 minutes ago, agent may be unhealthy
Fix:
kubectl logs -n rubixkube deployment/rubixkube-observer
kubectl rollout restart deployment/rubixkube-observer -n rubixkube

​
RubixKube takes 1-2 minutes to detect and analyze issues.

If failures are immediate but no insights after 5 minutes:
  • Check observer agent is running
  • Verify namespace is being watched (default: all namespaces)
  • Check RubixKube dashboard for agent status

​
What’s Next?

You’ve completed the basics! Now it’s time to see RubixKube detect real issues:

Continue to: Detecting Pod Failures

Learn by watching RubixKube detect and analyze common Kubernetes failures

​
Need Help?

Support

Email: [email protected]

Documentation

Browse our complete guides

Concepts

Understand how RubixKube works

Community

Join our Slack community