SysOM-AI: Continuous Cross-Layer Performance Diagnosis for Production AI Training

TL;DR

SysOM-AI is a continuous, low-overhead cross-layer observability system for diagnosing performance issues in large-scale AI training, significantly reducing diagnosis time.

Contribution

It introduces a novel system combining OS-level instrumentation and layered diagnosis for real-time performance troubleshooting in production AI training.

Findings

Diagnosed 94 production issues at Alibaba over a year.

Reduced median diagnosis time from days to about 10 minutes.

Achieved less than 0.4% overhead in continuous monitoring.

Abstract

Performance diagnosis in production-scale AI training is challenging because subtle OS-level issues can trigger cascading GPU delays and network slowdowns, degrading training efficiency across thousands of GPUs. Existing profiling tools are limited to single system layers, incur prohibitive overhead (10--30%), or lack continuous deployment capabilities, resulting in manual analyses spanning days. We argue that continuous, cross-layer observability enabled by OS-level instrumentation and layered differential diagnosis is necessary to address this gap. We introduce SysOM-AI, a production observability system that continuously integrates CPU stack profiling, GPU kernel tracing, and NCCL event instrumentation via adaptive hybrid stack unwinding and eBPF-based tracing, incurring less than 0.4% overhead. Deployed at Alibaba across over 80,000 GPUs for more than one year, SysOM-AI helped diagnose 94 confirmed production issues, reducing median diagnosis time from days to approximately 10 minutes.