Skip to main content

Server Activity Tracking

Monitor detailed activity and performance metrics for individual MCP servers.

Overview​

The Server Activity page provides deep insights into individual server performance, tool usage, and execution history.

Accessing Server Activity​

  1. Navigate to Servers page
  2. Click on a server
  3. Click "View Activity" button

Or access directly via URL: /servers/{id}/activity

Activity Dashboard​

Performance Metrics​

  • Total Tool Calls: Lifetime call count
  • Average Latency: Mean response time
  • Error Rate: Percentage of failed calls
  • Last Activity: When last used

Tool Usage Breakdown​

Pie chart showing:

  • Distribution of tool calls
  • Most frequently used tools
  • Tool-specific metrics

Timeline Chart​

24-hour activity timeline:

  • Request volume over time
  • Error spikes
  • Usage patterns

Recent Events​

List of recent tool calls:

  • Timestamp
  • Tool name
  • Latency
  • Success/failure status
  • Agent that made the call

Real-time Updates​

Activity page updates in real-time via WebSocket:

  • New tool calls appear immediately
  • Metrics update live
  • Charts refresh automatically

Filtering​

Filter events by:

  • Time Range: Last hour, 24 hours, week
  • Tool Name: Specific tools
  • Status: Success or failure only
  • Agent: Calls from specific agents

Server Types​

STDIO Servers​

Shows:

  • Tool call metrics
  • STDIO communication logs
  • Process status

HTTP Servers​

Shows:

  • HTTP request metrics
  • Response codes
  • Proxy latency
  • SSE stream activity

Best Practices​

  1. Monitor regularly: Check high-traffic servers frequently
  2. Watch error rates: Investigate spikes in failures
  3. Analyze patterns: Identify peak usage times
  4. Compare latency: Track performance trends

For API access, see Metrics API.