In this paper we show that the data plane of commodity programmable Network Interface Cards (NICs) can run neural network inference tasks required by packet monitoring applications, with low overhead. This is particularly important as the data transfer costs to the host system and dedicated machine learning accelerators, e.g., GPUs, can be more expensive than the processing task itself. We design and implement our system -N3IC -on two different NICs and we show that it can greatly benefit three different network monitoring use cases that require machine learning inference as first-class-primitive. N3IC can perform inference for millions of network flows per second, while forwarding traffic at 40Gb/s. Compared to an equivalent solution implemented on a general purpose CPU, N3IC can provide 100x lower processing latency, with 1.5x increase in throughput.

The capacity and programmability of reconfigurable hardware such as FPGAs has improved steadily over the years, but they do not readily provide any mechanisms for monitoring or debugging running programs. Such mechanisms need to be written into the program itself. This is done using ad hoc methods and primitive tools when compared to CPU programming. This complicates the programming and debugging of reconfigurable hardware.We introduce Program-hosted Directability (PhD), the extension of programs to interpret direction commands at runtime to enable debugging, monitoring and profiling. Normally in hardware development such features are fixed at compile time. We present a language of directing commands, specify its semantics in terms of a simple controller that is embedded with programs, and implement a prototype for directing network programs running in hardware. We show that this approach affords significant flexibility with low impact on hardware utilisation and performance.

No abstract

The idea to enable advanced in-network monitoring functionality has been lately fostered by the advent of massive data-plane programmability. A specific example includes the detection of traffic aggregates with programmable switches, i.e., heavy hitters. So far, proposed solutions implement the mining process by partitioning the network stream in disjoint windows. This practice allows efficient implementations but comes at a well-known cost: the results are tightly coupled with the traffic and window's characteristics. This poster quantifies the limitations of disjoint time windows approaches by showing that they hardly cope with traffic dynamics. We report the results of our analysis and unveil that up to 34% of the total number of the hierarchical heavy hitters might not be detected with those approaches. This is a call for a new set of windowless-based algorithms to be implemented with the match-action paradigm.

No abstract