CI Jobs Fixed, Created, or Deleted - January 2026 This document tracks all CI jobs that were fixed, created, modified, or deleted through dims' PRs in January 2026.
Status
Count
Created (New)
1
Fixed
12
Modified/Improved
8
Deleted
12
Total Jobs Affected 33
Job Name
PRs
Description
Prow
TestGrid
ci-kubernetes-e2e-kind-audit #36201 , #36202 , #36204 , #36205 , #36206 , #36212 , #36216 , #36225 New periodic job for APISnoop coverage. Mirrors ci-kubernetes-e2e-gci-gce but on KIND with audit logging. Required 8 PRs to get fully working (cluster creation fix, FOCUS handling, Serial test enablement, kubelet startup fix). Also required kind#4094 .
Prow TestGrid
Job Name
PRs
Issue
Fix Description
Prow
TestGrid
ci-aws-ec2-janitor #36324 --sweep-count flag not supportedRemoved unsupported --sweep-count=2 flag that was accidentally added. The aws-janitor image doesn't support this flag.
Prow TestGrid
pull-kubernetes-integration-go-compatibility #36159 Running incorrectly on master
Fixed branch targeting to prevent job from running on master branch where it shouldn't.
Prow TestGrid
pull-kubernetes-node-kubelet-serial-containerd #36304 Not triggering on relevant changes
Added run_if_changed trigger for test/e2e_node/ to catch issues early. This is a release-blocking job that was failing unnoticed.
Prow TestGrid
ci-cos-containerd-e2e-ubuntu-gce #36332 cgroupv1 not supported in k8s 1.36
Switched from ubuntu-os-gke-cloud to ubuntu-os-cloud for cgroupv2 support. Kubelet was failing with "cgroup v1 support is unsupported".
Prow TestGrid
ci-kubernetes-local-e2e #36333 Intermittent networking failures
Moved to eks-prow-build-cluster to investigate docker-in-docker networking issues.
Prow TestGrid
pull-kubernetes-local-e2e #36333 Intermittent networking failures
Moved to eks-prow-build-cluster (same as periodic).
Prow TestGrid
metrics-bigquery #36269 , #36272 Stale jobs in output + YAML null handling crash
Added stale job filtering and fixed crash when YAML has explicit null values (e.g., postsubmits: null).
Prow N/A
pull-kubernetes-e2e-gce-100-performance #36322 Outdated containerd/runc
Updated to containerd v2.1.5 and runc v1.3.4.
Prow TestGrid
ci-kubernetes-e2e-gci-gce-alpha-enabled-default #36322 Outdated containerd/runc
Updated to containerd v2.1.5 and runc v1.3.4.
Prow TestGrid
pull-e2e-gci-gce-alpha-enabled-default #36322 Outdated containerd/runc
Updated to containerd v2.1.5 and runc v1.3.4.
Prow TestGrid
pull-kubernetes-e2e-nftables-presubmit #36313 ImageVolume tests failing
Upgraded containerd from 1.7.24 (cos-113) to 1.7.27 for ImageVolume CRI extension support.
Prow TestGrid
ci-kubernetes-e2e-nftables #36313 ImageVolume tests failing
Upgraded containerd for ImageVolume support (same as presubmit).
Prow TestGrid
Job Name
PRs
Change
Prow
TestGrid
ci-kubernetes-e2e-ubuntu-gce-containerd #36128 Upgraded Ubuntu 22.04 → 24.04 LTS
Prow TestGrid
ci-kubernetes-e2e-gci-gce-serial-containerd #36128 Upgraded Ubuntu 22.04 → 24.04 LTS
Prow TestGrid
ci-kubernetes-e2e-gci-gce-serial-containerd-flaky #36128 Upgraded Ubuntu 22.04 → 24.04 LTS
Prow TestGrid
ci-kubernetes-e2e-serial-gce #36128 Upgraded Ubuntu 22.04 → 24.04 LTS
Prow TestGrid
ci-kubernetes-e2e-gce-ingress-lb-controller #36128 Upgraded Ubuntu 22.04 → 24.04 LTS
Prow TestGrid
post-kubernetes-push-e2e-glibc-dns-testing #36276 Renamed from jessie-dnsutils to glibc-dns-testing
Prow N/A
post-kubernetes-push-e2e-pytorch-wide-deep #36279 Replaced TensorFlow with PyTorch for node perf tests
Prow N/A
Minikube jobs (23 total) #36306 Standardized naming: pull-minikube-, ci-minikube- , post-minikube-*
Prow (minikube) TestGrid
Job NamePRs
Reason
Last Prow Link
ci-kubernetes-e2e-gce-network-metric-measurement #36163 Persistently failing, no longer maintained
Prow (archived)
post-kubernetes-push-e2e-cuda-vector-add #36256 Image removed in kubernetes#127141
N/A
post-kubernetes-push-e2e-cuda-vector-add-old #36256 Image removed in kubernetes#127141
N/A
post-kubernetes-push-e2e-echoserver #36256 Orphaned - image directory deleted
N/A
post-kubernetes-push-e2e-fakegitserver #36256 Orphaned - image directory deleted
N/A
post-kubernetes-push-e2e-iperf #36256 Orphaned - image directory deleted
N/A
post-kubernetes-push-e2e-jessie-dnsutils #36256 Replaced by glibc-dns-testing
N/A
post-kubernetes-push-e2e-logs-generator #36256 Orphaned - image directory deleted
N/A
post-kubernetes-push-e2e-net #36256 Orphaned - image directory deleted
N/A
post-kubernetes-push-e2e-porter #36256 Orphaned - image directory deleted
N/A
post-kubernetes-push-e2e-serve-hostname #36256 Orphaned - image directory deleted
N/A
post-kubernetes-push-e2e-tf-wide-deep #36279 Replaced by pytorch-wide-deep
N/A
Test Fixes in kubernetes/kubernetes These PRs fixed flaky or failing tests that were impacting CI jobs:
Test/Area
PRs
Issue
Fix
KubeProxy CLOSE_WAIT test #136554 , #136529 Failing in IPv6 environments, missing /proc/net/nf_conntrack
Read /proc/net/nf_conntrack directly instead of conntrack binary; handle IPv6 and missing file
ImagePullTest #136486 Infinite hang in poll loop
Added timeout to prevent hangs
container_restart_test #136484 AfterEach failing when CRI Proxy undefined
Added nil check for CRI Proxy
volume/nfs tests #136445 Failing on CentOS Stream 9
Fixed NFS image compatibility
Node perf tests #136398 TensorFlow build broken
Switched to PyTorch Wide-Deep
DNS tests #136358 , #136359 jessie-dnsutils image EOL
Created glibc-dns-testing replacement
Kops Scalability Job Improvements
Job Area
PRs
Change
Kops scalability tests kops#17891 , kops#17892 Added EtcdEventsHTTP feature flag and enabled it in scalability tests. Reduces TLS handshake overhead for events etcd, matching GCE scale test behavior.
Generated: January 29, 2026
