Incident Library

Root cause

Workload imbalance

작업 불균형 · workload-imbalance

1 case · 2025

Component Language runtime settings
Contributing factor Unreviewed defaults
Symptom Throughput degradationLoad imbalance
Organisation ByteDance

Cases

  1. IL-0417 Stragglers slowed 42.5% of large LLM training jobs on ByteDance's cluster, mostly without any faulty serverByteDance · 2025

View on the graph