Skip to content

Dashboard Metrics Reference

Complete reference for admin dashboard metrics, health indicators, and activity tracking.

Admin Dashboard


Definition: Number of users currently logged in with active sessions

Data Type: Integer (real-time count)

Calculation: Count of sessions with last_activity within the past 15 minutes

Use Cases:

  • Monitoring concurrent usage
  • Capacity planning
  • Detecting unusual spikes

Definition: Total number of properties in the database

Data Type: Integer

Calculation: Count of all records in the properties table

Includes:

  • Active listings
  • Pending listings
  • Sold properties (if retention policy includes them)

Excludes:

  • Soft-deleted properties
  • Draft listings (if applicable)

Definition: Status of external service integrations

Possible Values:

  • Healthy — All services responding normally
  • Degraded — Some services experiencing issues
  • Down — Critical service unavailable

Monitored Services:

  • MLSGrid API endpoint
  • Ollama vision server
  • Keycloak authentication
  • Database connection

Health Check Frequency: Every 60 seconds


Definition: Freshness of MLS data

Display Format: “Last synced X hours ago” or timestamp

Data Type: DateTime (last successful sync)

Color Coding:

  • Fresh — Synced within 6 hours
  • Stale — Synced 6-24 hours ago
  • Critical — Synced >24 hours ago

Related Actions: Manual sync trigger button appears when status is stale


Monitors connectivity to the MLSGrid API.

Status Color Description
Green Success Active sync, successful API responses
Yellow Warning Delayed responses, intermittent failures
Red Error Connection timeout, authentication failure

Diagnostic Data:

  • Last successful API call timestamp
  • Current error rate (if any)
  • Response time average (ms)

Monitors the semantic search embedding system.

Status Color Description
Green Success Embeddings up-to-date, vision extraction running
Yellow Warning Embeddings outdated, extraction pending
Red Error Ollama unreachable, no embeddings generated

Diagnostic Data:

  • Percentage of properties with embeddings
  • Last vision extraction job timestamp
  • Pending extraction queue size

Monitors external integrations and dependencies.

Status Color Description
Green Success All integrations responding normally
Yellow Warning Non-critical service degraded (e.g., analytics)
Red Error Critical service down (e.g., authentication)

Monitored Services:

  • Keycloak (authentication)
  • Ollama (embeddings)
  • External APIs (if configured)

Monitors disk usage for database and media storage.

Status Color Description
Green Success <70% usage
Yellow Warning 70-90% usage
Red Error >90% usage

Diagnostic Data:

  • Total disk space
  • Used space
  • Available space
  • Percentage used

Data Captured:

  • User ID and email
  • Login timestamp (ISO 8601)
  • IP address
  • User agent
  • Authentication method (OIDC, API key)

Display Format:

John Doe (john@example.com) logged in
2024-12-28 10:30:15 from 203.0.113.42

Data Captured:

  • User ID and email
  • Logout timestamp
  • Session duration
  • Logout type (manual, timeout, forced)

Display Format:

Jane Smith (jane@example.com) logged out
Session duration: 2h 15m

Data Captured:

  • Entity type (property, agent, brokerage, etc.)
  • Action (created, updated, deleted)
  • Entity ID
  • User who made the change
  • Timestamp
  • Changed fields (summary)

Display Format:

Sarah Admin updated Property #123456
Changed: price ($500,000 → $485,000)
2024-12-28 14:22:00

Data Captured:

  • Endpoint path
  • HTTP method
  • Response status code
  • Response time (ms)
  • API key ID (if authenticated)
  • Timestamp

Display Format:

GET /api/v1/properties → 200 OK (45ms)
API Key: CRM Integration
2024-12-28 16:45:30

Data Captured:

  • Error type (4xx, 5xx, application error)
  • Error message
  • Stack trace (if applicable)
  • User/API key associated
  • Request details
  • Timestamp

Display Format:

500 Internal Server Error
POST /api/v1/properties
Error: Database connection timeout
User: Admin (admin@example.com)
2024-12-28 18:10:00

Available filters in the dashboard:

Filter Description
Last Hour Events in past 60 minutes
Last 24 Hours Events in past day
Last 7 Days Events in past week
Custom Range User-specified start/end dates

Filter by event type:

  • All Activity — No filtering
  • User Events — Logins, logouts, profile changes
  • Data Changes — Create, update, delete operations
  • API Activity — External API calls
  • Errors Only — Failed requests, exceptions

Filter by importance:

Level Badge Examples
Info Info User login, data view
Warning Warning Rate limit approach, slow query
Error Error Failed API call, exception
Critical Critical Service down, data loss

How long different metrics are stored:

Metric Type Retention Period
Real-time Metrics 24 hours (raw data)
Aggregated Metrics 90 days (hourly)
Activity Logs 90 days
Error Logs 180 days
Audit Logs 1 year

Dashboard data can be exported in:

Format Use Case
CSV Spreadsheet analysis
JSON Programmatic processing
PDF Management reports