Understanding Data Storage and Ingestion for Large-Scale Deep Recommendation Model Training

Authors:
  • MZMark Zhao
    Mark Zhao
  • NANiket Agarwal
    Niket Agarwal
  • ABAarti Basant
    Aarti Basant
  • BGBugra Gedik
    Bugra Gedik
  • SPSatadru Pan
    Satadru Pan
  • MOMustafa Ozdal
    Mustafa Ozdal
  • RKRakesh Komuravelli
    Rakesh Komuravelli
  • JPJerry Pan
    Jerry Pan
  • TBTianshu Bao
    Tianshu Bao
  • HLHaowei Lu
    Haowei Lu
  • SNSundaram Narayanan
    Sundaram Narayanan
  • JLJack Langman
    Jack Langman

Abstract
Datacenter-scale AI training clusters consisting of thousands of domain-specific accelerators (DSA) are used to train increasingly-complex deep learning models. These clusters rely on a data storage and ingestion (DSI) pipeline, responsible for storing exabytes of training data and serving it at tens of terabytes per second. As DSAs continue to push training efficiency and throughput, the DSI pipeline is becoming the dominating factor that constrains the overall training performance and capacity. Innovations that improve the efficiency and performance of DSI systems and hardware are urgent, demanding a deep understanding of DSI characteristics and infrastructure at scale. This paper presents Meta's end-to-end DSI pipeline, composed of a central data warehouse built on distributed storage and a Data PreProcessing Service that scales to eliminate data stalls. We characterize how hundreds of models are collaboratively trained across geo-distributed datacenters via diverse and continuous training jobs. These training jobs read and heavily filter massive and evolving datasets, resulting in popular features and samples used across training jobs. We measure the intense network, memory, and compute resources required by each training job to preprocess samples during training. Finally, we synthesize key takeaways based on our production infrastructure characterization. These include identifying hardware bottlenecks, discussing opportunities for heterogeneous DSI hardware, motivating research in datacenter scheduling and benchmark datasets, and assimilating lessons learned in optimizing DSI infrastructure.