Issues

• Suspended TrainJobs created on v2.3.0 cannot resume after upgrading to v2.4.0

  area/controllerkind/bug
  kubeflow/trainer#4441 · Filter by author abhishekKokadwar opened on Nov 11, 2026
  2

  

• Torch/JAX/XGBoost runtime whose trainer job isn't named node is accepted but hangs

  kind/bugneeds-triage
  kubeflow/trainer#4422 · Filter by author JerT33 opened on Nov 8, 2026
  2

  

• TrainJob is failed immediately when activeDeadlineSeconds exceeds 9223372036

  kubeflow/trainer#4418 · Filter by author nikhil3495 opened on Nov 8, 2026
  1

• fix: cover TrainJob lookup failure in cache initializer

  kubeflow/trainer#4415 · Filter by author pratik-naik003 opened on Nov 7, 2026
  1

  

• OptimizationJob with DRA fields in trainJobTemplate is admitted while TrainJobDynamicResourceAllocation is disabled, then never creates a trial

  kubeflow/trainer#4413 · Filter by author harshrajdebug opened on Nov 6, 2026

• bug(runtime): Workload-Aware Scheduling gang minCount counts multi-slice Pods replicas times

  kubeflow/trainer#4411 · Filter by author singhharsh1708 opened on Nov 6, 2026
  1

• Fix the statusserver auth to use offline OIDC token validation using the API server's internal discovery endpoint

  kind/cleanup
  kubeflow/trainer#4399 · Filter by author robert-bell opened on Nov 4, 2026
  5

  

• bug(framework): PlainML plugin EnforceMLPolicy does not ignore Flux and XGBoost MLPolicySource

  kind/bugneeds-triage
  kubeflow/trainer#4397 · Filter by author iHamza14 opened on Nov 3, 2026
  3

  

• resourcesPerNode with only limits keeps the runtime's higher requests, so the trainer Pods are rejected

  kind/bugneeds-triage
  kubeflow/trainer#4390 · Filter by author aarushitandon0 opened on Nov 3, 2026
  1

  

• Forward AWS_* environment variables from DataCacheInitializer to LeaderWorkerSet containers

  kind/bugneeds-triage
  kubeflow/trainer#4388 · Filter by author Neha-Nimsatkar opened on Nov 3, 2026
  3

  

• TrainJob create is denied while the snapshot of a deleted TrainJob with the same name still exists

  kind/bug
  kubeflow/trainer#4372 · Filter by author Ahmad-Faraj opened on Nov 2, 2026
  1

  

• Discussion: Add a stable label to all Pods of a TrainJob

  area/apikind/discussionkind/featureneeds-triage
  kubeflow/trainer#4370 · Filter by author robert-bell opened on Oct 31, 2026
  2

  

• test: cover namespace resolution fallback in cache initializer

  kubeflow/trainer#4359 · Filter by author pratik-naik003 opened on Oct 30, 2026
  2

  

• Auto-merge Dependabot PRs when CI passes

  area/engprodgood first issuehelp wantedkind/feature
  kubeflow/trainer#4347 · Filter by author andreyvelich opened on Oct 28, 2026
  6

  

• A TrainJob with a negative resourcesPerNode request is admitted and crash-loops the Volcano scheduler

  kubeflow/trainer#4345 · Filter by author harshrajdebug opened on Oct 28, 2026

• Validate runtime patch volume references during EnforceMLPolicy stage

  kubeflow/trainer#4342 · Filter by author Prat260104 opened on Oct 28, 2026
  4

  

• TrainJob controller ignores groupKindConcurrency and always runs one worker

  kubeflow/trainer#4339 · Filter by author Ahmad-Faraj opened on Oct 27, 2026
  

• HuggingFace dataset initializer fails with IndexError for dataset ids without a namespace (hf://imdb)

  kubeflow/trainer#4315 · Filter by author Aryan-Pardeshi opened on Oct 26, 2026
  2

• S3 initializer matches ignore_patterns against the full object key, so relative patterns never match

  kubeflow/trainer#4313 · Filter by author Aryan-Pardeshi opened on Oct 26, 2026
  1

• ExtractResourcePerNodeFromRuntime panics on a replicatedJob whose Job template omits metadata, so TrainJob creation is rejected with an opaque 500

  kind/bugneeds-triage
  kubeflow/trainer#4303 · Filter by author aarushitandon0 opened on Oct 24, 2026
  1

  

• test(helm): add helm-unittest tests for the public ConfigMap

  kind/featureneeds-triage
  kubeflow/trainer#4299 · Filter by author shreyaabaranwal opened on Oct 23, 2026
  1

  

• Proposal: allow selecting ReplicatedJobs for PodGroupPolicy in sequential runtimes

  kubeflow/trainer#4298 · Filter by author sfeng1996 opened on Oct 23, 2026
  1

  

• TrainJob accepts env names and resource requests that its Pods cannot use, and then never starts

  kubeflow/trainer#4292 · Filter by author harshrajdebug opened on Oct 21, 2026

• OptimizationJob accepts trainer env vars its trials can't use, then stalls without failing

  kubeflow/trainer#4285 · Filter by author rshdhere opened on Oct 21, 2026
  1

  

• perf(cache): reuse one Flight connection per worker instead of reconnecting for every batch

  kubeflow/trainer#4280 · Filter by author DEVJHAWAR11 opened on Oct 20, 2026
  1