We supply Linux high availability clustering solutions. The aim is to use a set of building blocks to provide highly available and reliable mission critical applications.
We use a variety of technologies including disk replication, cluster resource management software and storage cluster technologies. These include a combination of Ceph, GlusterFS, Haproxy, DRBD, Corosync, Pacemaker , Keepalived and more.
The diagram below shows how you can use disk replication to create a Linux Disaster Recovery virtualization solution that replicates virtual machines in real time so that if a failure occurs, the synchronized virtual machine can just be booted off the second KVM hypervisor.
HAProxy Load Balancing
- Layer 4 (TCP) and Layer 7 (HTTP) load balancing
- Advanced health checking with customizable checks
- Support for sticky ses sions and persistence
- Detailed real-time monitoring and statistics
- SSL/TLS termination and offloading
- Dynamic configuration through runtime API
HAProxy can be deployed in various configurations:
- Active-Passive Setup
- Two HAProxy instances where one is active and the other is standby ○ Failover managed through Keepalived using VRRP protocol
- Ensures zero downtime during maintenance or failures
- Active-Active Setup
- Multiple active HAProxy instances sharing the load
- DNS round-robin or additional load balancer distributing traffic
- Maximum resource utilization and scalability
- Multi-tier Architecture
- Layer 7 proxying for web applications
- Layer 4 proxying for database and application servers
- Separate management of external and internal traffic
Traefik Modern Proxy
Traefik is a modern HTTP reverse proxy and load balancer designed for microservices and container environments. Its key features include:
- Automatic service discovery and configuration
- Native integration with Docker, Kubernetes, and other platforms
- Real-time configuration updates without restarts
- Let’s Encrypt integration for automatic SSL/TLS
- Advanced routing and middleware capabilities
- Modern dashboard with real-time metrics
Traefik deployment patterns:
- Container Orchestration
- Automatic discovery of Docker containers
- Dynamic routing based on container labels
- Seamless scaling and updates
- Kubernetes Integration
- Native Kubernetes Ingress controller
- Custom Resource Definitions (CRDs) support
- Automatic pod discovery and routing
- High Availability Setup
- Multiple Traefik instances for redundancy
- Shared configuration through key-value stores
- Integration with cluster management tools
Both HAProxy and Traefik can be integrated with our existing high availability stack:
- Corosync/Pacemaker for cluster management
- Keepalived for floating IP management
- DRBD for configuration replication
- Monitoring through Prometheus and Grafana
These proxies provide critical load balancing and routing capabilities essential for building robust, highly available infrastructure.