Homunculus: Auto-Generating Efficient Data-Plane ML Pipelines for Datacenter Networks
Homunculus: Auto-Generating Efficient Data-Plane ML Pipelines for Datacenter Networks
复制标题
DOI：
10.1145/3582016.3582022
复制
发表时间：
2022-06
期刊：
影响因子：
--
通讯作者：
Tushar Swamy;Annus Zulfiqar;Luigi Nardi;M. Shahbaz;K. Olukotun
中科院分区：
文献类型：
--
作者：
Tushar Swamy;Annus Zulfiqar;Luigi Nardi;M. Shahbaz;K. Olukotun
在过去十年中，网络中对机器学习（ML）应用程序的支持有了显着改善。公共数据集和可编程交换结构（包括对其进行编程的低级语言）的可用性为程序员提供了部署网络内ML的全栈。然而，所涉及的工具的多样性，加上ML模型设计和超参数调整的复杂优化任务，同时遵守网络约束（如吞吐量和延迟），使网络运营商有责任成为ML，网络设计和可编程硬件方面的专家。我们提出了Homunculus，这是一个高级框架，使网络运营商能够以声明式而不是命令式的方式指定他们的ML需求。Homunculus将训练数据以及伴随的网络和硬件约束作为输入，并自动生成合适的模型并将其安装到底层交换目标上。它执行模型设计-空间探索，训练和平台代码生成作为编译器阶段，让网络运营商专注于获取高质量的网络数据。我们对现实世界的ML应用程序的评估表明，与手动调整的替代方案相比，Homunculus生成的模型的F1得分提高了12%，同时在底层目标的资源限制内运行。我们进一步展示了在新兴的每包ML平台上生成的模型的高性能和更高的反应性（秒到纳秒），以展示Homunculus的及时性和实际意义。
Support for Machine Learning (ML) applications in networking has significantly improved over the last decade. The availability of public datasets and programmable switching fabrics (including low-level languages to program them) presents a full-stack to the programmer for deploying in-network ML. However, the diversity of tools involved, coupled with complex optimization tasks of ML model design and hyperparameter tuning while complying with the network constraints (like throughput and latency), puts the onus on the network operator to be an expert in ML, network design, and programmable hardware. We present Homunculus, a high-level framework that enables network operators to specify their ML requirements in a declarative rather than imperative way. Homunculus takes as input the training data and accompanying network and hardware constraints, and automatically generates and installs a suitable model onto the underlying switching target. It performs model design-space exploration, training, and platform code-generation as compiler stages, leaving network operators to focus on acquiring high-quality network data. Our evaluations on real-world ML applications show that Homunculus’s generated models achieve up to 12% better F1 scores compared to hand-tuned alternatives, while operating within the resource limits of the underlying targets. We further demonstrate the high performance and increased reactivity (seconds to nanoseconds) of the generated models on emerging per-packet ML platforms to showcase Homunculus’s timely and practical significance.