Managing Clusters
Create, configure, and monitor infrastructure clusters
A Cluster in FixPanic represents a logical grouping of computing resources where your agents run. Think of clusters as environments (like "Production" or "Staging") that organize your infrastructure.
What is a Cluster?
Clusters serve several purposes:
- Organization - Group agents by environment, project, or function
- Isolation - Keep production and development separate
- Credentials - Each cluster has unique credentials for agent deployment
- Monitoring - Track health and status at the cluster level
Infrastructure Flexibility
FixPanic is infrastructure-agnostic. A cluster can represent:
- A single VPS or VM
- A dedicated bare-metal server
- Multiple servers in a data center
- Cloud instances (AWS, GCP, Azure)
Creating a Cluster
- Navigate to the Clusters page from the sidebar
- Click the "Create Cluster" button
- Enter cluster details:
- Name: A friendly identifier (e.g., "Production-East", "Dev-Environment")
- Description: Optional notes about the cluster's purpose
- Click "Create"
- Save your credentials immediately:
- Agent ID: Unique identifier for this cluster
- API Key: Authentication key for agents
Warning: The API key is only shown once! Save it immediately. If lost, you'll need to regenerate it.
Cluster Health
Clusters have health indicators based on agent connectivity:
| Status | Meaning | Action |
|---|---|---|
| Healthy | All agents connected and responding | None needed |
| Unhealthy | No heartbeat received for 5+ minutes | Check agent status |
| Degraded | Some agents disconnected | Investigate affected agents |
Health Checks
Agents periodically send "heartbeats" to the FixPanic platform:
- Healthy: Heartbeat received within the last minute
- Unhealthy: No heartbeat for more than 5 minutes
Cluster Actions
From the Clusters List
Click the actions menu (three dots) on any cluster row:
| Action | Description |
|---|---|
| View Details | See full cluster information and agents |
| Health Check | Trigger an immediate connectivity test |
| Pause | Stop scheduling new agents (existing continue running) |
| Resume | Re-enable agent scheduling |
| Regenerate Key | Create a new API key (invalidates old one) |
| Delete | Remove the cluster and all data |
Pausing a Cluster
When you pause a cluster:
- No new agents can be deployed to it
- Existing agents continue running
- The cluster is marked as "Paused" in the UI
This is useful for:
- Maintenance windows
- Temporarily disabling a test environment
- Preparing for migration
Deleting a Cluster
Important: You must delete all agents in a cluster before you can delete the cluster itself.
To delete a cluster:
- Navigate to the cluster's detail page
- Delete all agents in the cluster (or move them to another cluster)
- Return to the cluster settings
- Click Delete Cluster and confirm
Cluster Detail View
Click on a cluster name to see its detail page:
Overview Tab
- Cluster health status
- Number of active agents
- Creation date
- Last activity
Agents Tab
- List of all agents in this cluster
- Quick actions for each agent
- Deploy new agent button
Settings Tab
- Edit cluster name and description
- View/regenerate credentials
- Delete cluster
Filtering Clusters
On the Clusters page, you can filter by:
| Filter | Options |
|---|---|
| Status | All, Healthy, Unhealthy, Paused |
| Search | Filter by name or ID |
| Sort | Name, Created Date, Agent Count |
Best Practices
Naming Conventions
Use clear, consistent naming:
prod-us-east- Production in US East regionstaging-main- Main staging environmentdev-team-alpha- Development for Team Alpha
Environment Separation
Tip: Always keep production and development clusters separate. This prevents accidental commands from affecting production systems.
Credential Management
- Never share API keys across clusters
- Rotate keys periodically
- Store keys securely (secrets manager, environment variables)
- Never commit keys to version control
Troubleshooting
Cluster Shows "Unhealthy"
- SSH into your server where the agent is installed
- Check if the agent process is running:
fixpanic agent status - Restart the agent if needed:
sudo systemctl restart fixpanic-connectivity-layer - Check network connectivity:
nc -zv socket.fixpanic.com 9000
Can't Create New Cluster
- Check your subscription plan for cluster limits
- Ensure you have sufficient credits
- Contact support if the issue persists
Next Steps
- Managing Agents - Deploy and manage agents within your clusters.
- First Deployment - Deploy your first agent to a cluster.