[{"id":"IL-0137","title":{"en":"Training initialisation on a 60-node NVIDIA B200 cluster took more than 8 hours instead of minutes because fragmented small random I/O from all nodes saturated the metadata service of the VAST Data NFS storage","ko":"60노드 B200 학습 클러스터에서 노드들의 작은 랜덤 I/O 가 VAST 스토리지의 메타데이터 서비스를 포화시켜 학습 초기화가 8시간 넘게 걸렸다"},"org":"Lablup Inc","org_slug":"lablup-inc","date":"2026-06-15","severity":"unknown","root_cause":"capacity","url":"https://benchlet.com/incidentlib/app/cases/IL-0137","shared_components":["shared-filesystem"]}]