See what's inside your queues — not just message counts.
Get Started · Features · Screenshots · API Docs · Contributing
Production breaks at 2 AM. Azure Portal shows 5,000 messages in Dead-Letter Queue but you can't read them — only counts and metadata. You manually sample messages one by one, spending hours on what should take minutes.
ServiceHub is a self-hosted web application that lets you browse, search, and analyze Azure Service Bus messages in real time — like email for your message queues.
| Capability | Azure Portal | ServiceHub |
|---|---|---|
| View message content | Count only | Full body + properties |
| Search messages | Not available | Full-text search |
| DLQ investigation | One at a time | Batch analysis + AI |
| Pattern detection | Not available | AI-powered clustering |
| Replay from DLQ | Not available | One-click replay |
| Access | Cloud portal | Any browser |
Enter your connection string and you're in. Supports Listen-only (read-only) or Manage (full access) policies.
Browse active messages across queues and topic subscriptions. See message previews, status badges, and metadata at a glance.
Click any message to inspect every detail. System properties show Message ID, enqueue time, TTL, sequence number, delivery count, and content type.
Full JSON body with syntax highlighting and copy button. Supports JSON, XML, and plain text formats.
AI-powered analysis detects patterns, anomalies, and error clusters across your messages. All processing happens in your browser — no data leaves your environment.
Deep forensic analysis showing message lifecycle, delivery attempts, and processing timeline.
View all custom application properties and correlation headers.
Investigate failed messages with DLQ reason, error description, and AI-powered remediation guidance. Replay messages back to the original queue with one click.
See why messages failed and replay them after fixing the root cause.
AI automatically categorizes DLQ failures and suggests remediation steps.
Browse messages from topic subscriptions with the same powerful inspection tools.
Floating action button provides quick access to Send Message, Generate Messages, Test DLQ, and Refresh All.
Send single messages to queues or topics for ad-hoc testing. Supports custom properties, content types, and advanced options.
Generate realistic test messages with built-in scenarios (Order Processing, Payment Gateway, Notification Service, and more). Configure volume and anomaly rate.
Persistent tracking and monitoring of dead-letter queue messages with trend chart, status breakdown, and category classification.
Drill into individual DLQ records with forensic timeline, replay history, and status tracking.
Define conditional replay rules with live statistics. Match messages by dead-letter reason, error description, entity name, content type, or body text.
Browse pre-built rule templates for common failure scenarios — transient errors, max delivery exceeded, expired messages, and more.
Build custom rules with field conditions, operators, actions, rate limiting, and target entity configuration.
Monitor API health, uptime, memory usage, thread count, GC collections, and server information.
Searchable help guide covering every feature, Azure Service Bus concepts, and a guided tour.
Find messages by any property — message ID, correlation ID, body content, custom headers.
Namespace browser with live message counts, queue/topic tree, and Quick Access panel.
Interactive API documentation with Scalar — test endpoints directly from the browser.
- Browse active and dead-letter queue messages side by side
- View full message body with JSON syntax highlighting
- Inspect system properties, custom headers, and application properties
- Real-time search across message content and properties
- Auto-refresh with configurable polling intervals
- Pattern detection — Identify error clusters across thousands of messages
- Anomaly identification — Flag unusual messages automatically
- Remediation suggestions — Actionable guidance for each failure type
- Client-side processing — All analysis runs in your browser; no data leaves your environment
- Persistent tracking — DLQ messages stored in local SQLite database
- Category classification — Auto-categorizes: Transient, MaxDelivery, Expired, DataQuality, Authorization
- Trend chart — 30-day DLQ trend visualization (New vs. Resolved)
- Instant scanning — "Scan Now" for immediate DLQ polling
- Export — Download DLQ data as JSON or CSV
- Status tracking — Active → Replayed → ReplayFailed → Resolved
- Conditional matching — Match messages by reason, error description, entity, delivery count, body text
- Operators — Contains, Equals, StartsWith, EndsWith, Regex, GreaterThan, LessThan, In
- Live statistics — Pending/Replayed/Success counts updated in real time
- Rate limiting — Max replays per hour to prevent overwhelming downstream services
- Batch replay — Replay all matching messages with one click
- Template gallery — Pre-built rules for common failure scenarios
- Circuit breaker — Auto-disables rules if success rate drops below threshold
- Send Message — Send single messages to queues or topics with custom properties
- Generate Messages — 6 built-in scenarios with configurable volume (30–200) and anomaly rate (0–50%)
- Test DLQ — Move test messages to dead-letter queue for testing DLQ workflows
- Tagged messages — All generated messages tagged with
ServiceHub-Generatedfor easy cleanup
- Read-only by default — Uses Azure SDK PeekMessagesAsync; messages are never removed
- Listen-only supported — Works with Listen permission for browse-only access
- Encrypted at rest — Connection strings encrypted with AES-GCM
- No external API calls — AI analysis runs entirely in the browser
- No data persistence — Messages displayed in-memory only
- Safe for production — Will not interfere with your consumers
git clone https://github.com/debdevops/servicehub.git
cd servicehub
./run.shThe script automatically:
- Installs .NET 10 SDK (if not present)
- Installs Node.js 18+ (if not present)
- Builds and starts both API and UI servers
Open http://localhost:3000 and connect with your Azure Service Bus connection string.
Auto-installed by run.sh:
- .NET 10.0 SDK
- Node.js 18+
You provide:
- Azure Service Bus connection string (Listen permission minimum)
For read-only browsing:
az servicebus namespace authorization-rule create \
--namespace-name <your-namespace> \
--resource-group <your-rg> \
--name servicehub \
--rights ListenFor full access (send, generate, replay):
az servicebus namespace authorization-rule create \
--namespace-name <your-namespace> \
--resource-group <your-rg> \
--name servicehub \
--rights Listen Send Manage| Service | URL |
|---|---|
| UI | http://localhost:3000 |
| API | http://localhost:5153 |
| API Docs | http://localhost:5153/scalar/v1 |
servicehub/
├── apps/web/ # React 18 + TypeScript + Vite frontend
│ └── src/
│ ├── components/ # UI components (messages, DLQ, rules, FAB)
│ ├── hooks/ # React Query hooks for API communication
│ ├── lib/ # API client, utilities, help content
│ └── pages/ # Page components (Messages, Connect, Rules, Health, Help)
│
├── services/api/ # ASP.NET Core 10 backend
│ └── src/
│ ├── ServiceHub.Api/ # REST controllers, middleware, auth
│ ├── ServiceHub.Core/ # Domain entities, DTOs, interfaces
│ ├── ServiceHub.Infrastructure/ # Azure SDK integration, SQLite
│ └── ServiceHub.Shared/ # Common types and utilities
│
├── scripts/ # Setup and utility scripts
└── run.sh # One-command startup
Tech Stack:
| Layer | Technology |
|---|---|
| Frontend | React 18, TypeScript, Tailwind CSS, TanStack Query |
| Backend | ASP.NET Core 10, Azure.Messaging.ServiceBus SDK |
| AI Analysis | Client-side pattern detection (no external API) |
| Database | SQLite (DLQ Intelligence), in-memory cache |
| API Docs | Scalar (OpenAPI) |
For detailed backend architecture, see services/api/ARCHITECTURE.md.
Problem: 5,000 orders stuck in DLQ at 2 AM.
With ServiceHub:
- Browse all 5,000 DLQ messages instantly
- AI detects 3 error patterns: Payment Timeout (40%), Invalid Address (35%), Duplicate Order (25%)
- Search for specific customer orders by correlation ID
- Create auto-replay rule for Payment Timeout — replay all 2,000 messages
Time saved: 6 hours → 45 minutes
Problem: Customer reports order never processed.
With ServiceHub:
- Search across message body and properties
- Find order in 3 seconds across 10,000 messages
- Review full message properties and delivery history
Time saved: 30 minutes → 30 seconds
Problem: Need to test error handling with 100 realistic failure scenarios.
With ServiceHub:
- Select Payment Gateway scenario in Message Generator
- Generate 100 messages with 30% anomaly rate
- Verify error handling and DLQ behavior
- Clean up test messages filtered by ServiceHub-Generated tag
Time saved: Manual testing → Automated in 2 minutes
| Permission Level | Capabilities |
|---|---|
| Listen only | Browse messages, inspect DLQ, search, AI insights, view health |
| Listen + Send | All above + replay from DLQ + send test messages |
| Manage | All above + generate messages, test DLQ, full management |
Tip: Create a dedicated
servicehubpolicy instead of usingRootManageSharedAccessKey.
ServiceHub exposes a REST API documented with Scalar (OpenAPI). Access interactive docs at:
http://localhost:5153/scalar/v1
Key endpoints:
GET /api/v1/namespaces— List connected namespacesGET /api/v1/namespaces/{id}/queues— List queues with message countsGET /api/v1/namespaces/{id}/topics— List topics with subscription countsGET /api/v1/namespaces/{id}/queues/{name}/messages— Browse messagesPOST /api/v1/namespaces/{id}/queues/{name}/messages— Send a messageGET /api/v1/dlq-history— DLQ Intelligence recordsGET /api/v1/replay-rules— Auto-replay rules
Q: Does ServiceHub remove messages from queues? No. ServiceHub uses PeekMessagesAsync which reads messages without removing them. Your consumers continue processing normally.
Q: Is it safe for production? Yes. ServiceHub only requires Listen permission and operates in read-only mode. Messages remain in queues for your actual consumers.
Q: How does AI analysis work? ServiceHub analyzes message content using heuristic algorithms entirely in your browser. No data leaves your environment.
Q: What about sensitive data? Messages are displayed only in your browser session — not persisted. Deploy ServiceHub in your private network and restrict access.
Q: Can I deploy to Azure App Service / Kubernetes? Yes. ServiceHub is a standard ASP.NET Core + React SPA. Containerize with Docker and deploy anywhere supporting .NET 10.
Q: Does it support topics with subscriptions? Yes. Browse messages from both queues and topic subscriptions independently.
We welcome contributions! Bug fixes, features, and documentation improvements.
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Make your changes with tests
- Commit and push
- Open a Pull Request
MIT License — see LICENSE for details.
ServiceHub — Because your Service Bus messages should not be invisible during incidents.
Built for DevOps, Platform, and SRE Engineers.