ZipLine Article Swipe
YOU?
·
· 2020
· Open Access
·
· DOI: https://doi.org/10.1145/3386367.3431302
· OA: W3107507029
Network appliances continue to offer novel opportunities to offload\nprocessing from computing nodes directly into the data plane. One popular\nconcern of network operators and their customers is to move data increasingly\nfaster. A common technique to increase data throughput is to compress it before\nits transmission. However, this requires compression of the data -- a time and\nenergy demanding pre-processing phase -- and decompression upon reception -- a\nsimilarly resource consuming operation. Moreover, if multiple nodes transfer\nsimilar data chunks across the network hop (e.g., a given pair of switches),\neach node effectively wastes resources by executing similar steps. This paper\nproposes ZipLine, an approach to design and implement (de)compression at line\nspeed leveraging the Tofino hardware platform which is programmable using the\nP4_16 language. We report on lessons learned while building the system and show\nthroughput, latency and compression measurements on synthetic and real-world\ntraces, showcasing the benefits and trade-offs of our design.\n
