Architecture
Diese Seite ist noch nicht in deiner Sprache verfügbar. Englische Seite aufrufen
STACKIT RabbitMQ is built on a managed RabbitMQ platform, providing a robust and scalable message broker solution for your applications.
Core Components
Section titled “Core Components”RabbitMQ Message Broker
Section titled “RabbitMQ Message Broker”RabbitMQ serves as the central message broker, enabling asynchronous communication between applications and services. It supports multiple messaging protocols and patterns, including publish/subscribe, request/reply, and point-to-point messaging. The broker handles message routing, queuing, and delivery, ensuring reliable message processing across distributed systems.
Management Interface
Section titled “Management Interface”The RabbitMQ Management interface provides a web-based UI for monitoring and managing your message broker. It allows you to view queues, exchanges, bindings, and message rates, as well as manage users and permissions.
Plugin System
Section titled “Plugin System”RabbitMQ supports various plugins that extend its functionality, including federation, shovel, MQTT, STOMP, and consistent hash exchange plugins. These can be enabled and configured based on your specific use case requirements.
Infrastructure architecture
Section titled “Infrastructure architecture”Service instance isolation
Section titled “Service instance isolation”Each RabbitMQ service instance runs on dedicated virtual machines in the STACKIT Compute Engine. This ensures isolation from other services and protects against resource contention (bad neighborhood protection).
Deployment separation
Section titled “Deployment separation”RabbitMQ services created through Cloud Foundry and those created via STACKIT Portal/CLI are completely separated. Services created in one environment must be managed and deleted in the same environment - cross-environment management is not supported for technical reasons.
Single Node vs. Cluster
Section titled “Single Node vs. Cluster”Single Node
- Runs on a single VM
- Suitable for development and testing
- Not recommended for production use
Cluster (Replica Set)
- Consists of 3 nodes distributed across multiple availability zones
- Provides high availability and fault tolerance
- Recommended for production workloads
- Ensures service continuity even if a node or availability zone fails
- Automatic recovery in the event of equipment or data center failures
Network access
Section titled “Network access”RabbitMQ instances are accessible via public IP addresses, allowing integration with:
- Cloud Foundry applications
- Kubernetes workloads
- Virtual machines
- External systems
Access is secured through ACLs (Access Control Lists) and instance-level authentication. All connections support SSL/TLS encryption with up-to-date TLS v1.3 for enhanced security.
Data management
Section titled “Data management”Backup system
Section titled “Backup system”The integrated Backup Manager automatically creates full backups every 4 hours and retains them for 14 days. You can also trigger manual backups and restore from existing backups through the service dashboard.
Scalability
Section titled “Scalability”RabbitMQ uses a t-shirt sizing model with predefined service plans. When your workload grows, you can upgrade to a larger plan with more resources (vCPU, RAM, disk space). All instances utilize high-performance SSD storage for optimal performance.