The sniffer is engineered to maximize multi-core CPU utilization, capable of processing 10 Gbit VoIP traffic on a single server. Tested and running in production environments with over 100,000 concurrent calls.
Multi-Core Performance
- 10 Gbit Processing - Full line-rate VoIP traffic analysis on a single server with multi-core optimization
- 100,000+ Concurrent Calls - Tested and deployed in carrier-grade environments with massive call volumes
- Standard 10Gbit NICs - Works with commodity Intel network cards using native Linux capture or DPDK kernel-bypass
- DPDK Acceleration - Optional kernel-bypass packet capture for maximum throughput when needed
- Napatech Support - Hardware timestamping NICs supported, though not required thanks to DPDK efficiency
- ARM Architecture - Native support for ARM-based servers and embedded devices
Horizontal Scaling
- Distributed Architecture - Scale by deploying multiple independent sniffers writing to a central database
- Central GUI - Single web interface aggregates data from all distributed capture nodes
- 50,000 CDR Writes/Second - Database model optimized for extreme write throughput
- Serialized Packet Storage - Optimized format minimizes storage costs while maintaining full packet fidelity, ideal for cost-effective storage media
Flexible Deployment
- Port Mirroring - Passive capture via SPAN, RSPAN, or ERSPAN from network switches
- Direct Analysis - Deploy directly on SBC, PBX, or media gateways for integrated monitoring
- Virtual Machines - Run in VMware, KVM, Hyper-V, or any hypervisor environment
- Cloud Native - Full support for AWS, Google Cloud, Microsoft Azure, and all major cloud platforms
- Hybrid Deployments - Combine capture methods across on-premise and cloud infrastructure
Proven at Scale
Used by carriers and large enterprises worldwide. Our architecture has been refined over years of production deployments handling billions of calls with ultra-thin resource footprint.