Modern telco operators increasingly require continuous active performance monitoring across thousands or even tens of thousands of access nodes. While TWAMP has become the de facto standard for active IP performance measurements, implementing it at carrier scale is far from trivial.
This presentation explains how a carrier-scale TWAMP monitoring platform can be built using commodity x86 servers together with DPDK-based userspace packet processing that completely bypasses the Linux networking stack.
The talk discusses the architectural decisions required to support massive numbers of concurrent TWAMP sessions, including packet processing, session scheduling, timer management, CPU affinity, NUMA awareness and resource optimization.
Real benchmark results and operational experience from large-scale deployments demonstrate the achievable scalability and identify the practical limitations of this approach.
Attendees will gain practical insight into designing highly scalable active monitoring systems without relying on proprietary hardware appliances.
**Why carrier-scale TWAMP is difficult**
Most existing TWAMP implementations are designed for relatively small deployments. Scaling active measurements to tens of thousands of simultaneously monitored access nodes introduces significant challenges in packet processing, timing accuracy, scheduling and resource management.
**Why DPDK**
The presentation explains why bypassing the Linux kernel networking stack is essential for achieving deterministic latency and high packet processing rates. It discusses how DPDK enables direct NIC access, reduces interrupt overhead and provides predictable packet forwarding performance.
**System architecture**
Topics include:
- Userspace packet processing architecture
- DPDK forwarding engine and application-to-engine communication
- TWAMP session management
- Timestamping offload and packet pacing
- Multi-core scalability on commodity x86 hardware
**Performance**
Real benchmark results will be presented, including:
- maximum number of concurrent TWAMP sessions
- packets per second
- CPU utilisation
- memory consumption
- scaling across multiple CPU cores
- practical deployment limits
**Operational lessons learned**
The presentation concludes with practical experience gained from deploying large-scale active monitoring systems in carrier environments, including common pitfalls, optimisation techniques and recommendations for future deployments.
Licensed to the public under http://creativecommons.org/licenses/by/4.0
about this event: https://cfp.denog.de/denog18/talk/39VKX7/