cd /news/ai-infrastructure/vast-data-solves-job-level-noisy-nei… · home › topics › ai-infrastructure › article
[ARTICLE · art-171833] src=blocksandfiles.com ↗ pub= topic=ai-infrastructure verified=true sentiment=↑ positive

VAST Data solves job-level noisy neighbor problem

VAST Data introduced Job-Level QoS, a feature that lets administrators enforce per-job bandwidth and IOPS policies in its storage system by extending job identities from workload managers such as Slurm, PBS Pro and LSF into the VAST data layer. VAST Principal Technical Strategist Glenn Lockwood said the storage system "can now manage the I/O performance it delivers to individual jobs competing for resources, allowing resource managers and schedulers to prioritize bandwidth and IOPS without having to cancel jobs or change how applications run." Job-Level QoS is planned for general availability with VAST AI OS 5.6 in late January or early February 2027 and requires VAST NFS 4.8 or later; VAST will demonstrate the feature at SC26.

key takeaway

VAST AI OS 5.6 will let administrators set independent per-job bandwidth and IOPS policies via job identifiers carried in VAST NFS 4.8 client requests, adjustable in real time without restarting jobs, with telemetry through the open-source VNFS Collector into Prometheus, Grafana or the VAST DataBase

by read2 min views2 publishedNov 18, 2026
VAST Data solves job-level noisy neighbor problem
Image: Blocksandfiles (auto-discovered)

FLASH

VAST Data has made job level bandwidth and IOPS levels controllable so that longer-running jobs aren’t slowed by other jobs hogging bandwidth or IOPS.

The company says that administrators of large-scale computing environments have faced a familiar problem: a single I/O-intensive job can slow down dozens of others, with few options to intervene without disrupting ongoing work. On the world’s largest computing systems, a critical simulation can be delayed by an unrelated analysis workload, even when ample computing power is available. For research institutions and AI infrastructure providers, this competition creates unpredictable performance, operational disruption, and inefficient use of expensive infrastructure.

Glenn Lockwood, Principal Technical Strategist at VAST Data, said: “Anyone who's used a supercomputer knows the frustration of unpredictable performance when another job consumes more than its share of shared resources. The scheduler knows exactly which jobs are running, but that understanding has historically stopped at the storage system. With Job-Level QoS, we’re closing that gap. The storage system itself can now manage the I/O performance it delivers to individual jobs competing for resources, allowing resource managers and schedulers to prioritize bandwidth and IOPS without having to cancel jobs or change how applications run.”

Job-Level QoS extends the job identity defined by workload managers such as Slurm, PBS Pro, and LSF into the VAST data layer. This allows organizations to treat I/O performance as an allocatable resource alongside compute, enforcing individual bandwidth and IOPS policies on a job-by-job basis and adjusting allocations in real time while workloads continue running. Administrators can protect priority workloads and make more productive use of shared infrastructure without interrupting the research and applications depending on it.

VAST AI OS has several new controls to enable this:

  • Enforce QoS at the individual job level: The VAST NFS client carries configurable job identifiers with I/O requests, enabling VAST to apply independent bandwidth and IOPS policies to multiple jobs sharing a single compute node, or manage the collective I/O of a job spanning multiple nodes under a single policy.
  • Adjust performance without interruption: Administrators can modify read and write performance policies in real time, without restarting jobs, modifying applications, or remounting storage.
  • See which jobs are consuming resources: The optional, open-source VNFS Collector provides job-level visibility into bandwidth, IOPS, and errors, with telemetry available through Prometheus, Grafana, or the VAST DataBase.
  • Extend controls beyond batch: The underlying architecture also supports Kubernetes and other orchestrators, enabling workload-level performance management across AI training, inference, and scientific computing environments.

VAST will demonstrate Job-Level QoS in action at SC26,  showing how administrators can identify competing jobs, dynamically adjust bandwidth allocations, and enforce independent performance policies across shared infrastructure while workloads continue running uninterrupted.

Job-Level QoS is planned for general availability with VAST AI OS 5.6 in late January or early February 2027, requiring VAST NFS 4.8 or later.

── more in #ai-infrastructure 4 stories · sorted by recency
── more on @vast data 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/vast-data-solves-job…] indexed:0 read:2min 2026-11-18 · —