v2.6.0-rc.0

What's Changed

Kubecost 2.6 implements a new RBAC design. RC.0 is being tested and should not be deployed in production environments without assistance from Kubecost support.

Helm changes:

• Bump prometheus/prometheus from v3.0.0 to v3.0.1 in /cost-analyzer by @dependabot in #3757
• add helm config option to mount ca certs to cost model container by @mittal-ishaan in #3760
• correct settting log level in readme by @mittal-ishaan in #3731
• bump to kiwigrid 1.28.1 by @jessegoodier in #3764
• add init container for updating ca trust and shift getting ca cert from secret to config map by @mittal-ishaan in #3763
• bump cluster-controller by @cliffcolvin in #3773
• Remove-old-api-port by @jessegoodier in #3781
• Bump helm/kind-action from 1.10.0 to 1.11.0 by @dependabot in #3778
• Bump prometheus-operator/prometheus-config-reloader from v0.78.2 to v0.79.1 in /cost-analyzer by @dependabot in #3782
• Add SCC for clusterController on OpenShift by @ivankube in #3730
• Bump grafana/grafana from 11.3.1 to 11.4.0 in /cost-analyzer by @dependabot in #3768
• Bump prometheus-operator/prometheus-config-reloader from v0.79.1 to v0.79.2 in /cost-analyzer by @dependabot in #3790
• remove extra tab spaces in Helm values file in openshift in-cluster prom configuration by @mittal-ishaan in #3793
• Update ca cert in cloud cost and aggregator container too by @mittal-ishaan in #3794
• bump cluster controller by @cliffcolvin in #3795
• Bump helm/kind-action from 1.11.0 to 1.12.0 by @dependabot in #3792
• Bump kiwigrid/k8s-sidecar from 1.28.1 to 1.28.4 in /cost-analyzer by @dependabot in #3791
• bump modeling 0.1.19 fix cve-2024-12254 by @cliffcolvin in #3775
• Bump prometheus/prometheus from v3.0.1 to v3.1.0 in /cost-analyzer by @dependabot in #3800
• remove unused timezone offset values by @jessegoodier in #3809
• Bump kiwigrid/k8s-sidecar from 1.28.4 to 1.29.0 in /cost-analyzer by @dependabot in #3806
• teams rewrite secret/env functionality by @kaelanspatel in #3799
• fix links for linter by @jessegoodier in #3820
• Route /model auth endpoints to the aggregator by @kaelanspatel in #3816
• Update values.yaml to define teams configuration, as opposed to commented out by @thomasvn in #3821
• Bump prometheus/alertmanager from v0.27.0 to v0.28.0 in /cost-analyzer by @dependabot in #3815
• Bump helm/chart-testing-action from 2.6.1 to 2.7.0 by @dependabot in #3814
• fix: add missing url in grafana datasource by @takirala in #3812
• fix: nginx read only file issue for ubi images for frontend images by @mittal-ishaan in #3810
• Bump kiwigrid/k8s-sidecar from 1.29.0 to 1.29.1 in /cost-analyzer by @dependabot in #3823
• Bump in-code version of v2.6 branch for 2.6.0-rc.0 by @cliffcolvin in #3824

New Contributors

• @takirala made their first contribution in #3812

Full Changelog: v2.5.3...v2.6.0-rc.0

v2.5.3

What's Changed

• Fixed Allocation API for ramByteRequestAverage, ramByteUsageAverage, cpuCoreRequestAverage, cpuCoreUsageAverage, gpuUsageAverage, and gpuRequestAverage fields. Their response values now align with the behavior in v2.0, v2.1, and v2.2. When aggregated (e.g., across a namespace), the values represent the SUM of all containers, not the AVG.
• Fixed ui/Kai oh uh cloud cost bug.

v2.5.3-rc.1

What's Changed

• Fixed Allocation API for ramByteRequestAverage, ramByteUsageAverage, cpuCoreRequestAverage, cpuCoreUsageAverage, gpuUsageAverage, and gpuRequestAverage fields. Their response values now align with the behavior in v2.0, v2.1, and v2.2. When aggregated (e.g., across a namespace), the values represent the SUM of all containers, not the AVG.

v2.5.3-rc.0

What's Changed

• Fix for ui/Kai Uh Oh! error on cloud costs page

v2.5.2

What's Changed

• Fix cloud cost filter consistency
• Fix container request-sizing pod name display
• Fix container request-sizing csv download to export all rows
• Fix GPU savings showing idle load balancer as a GPU resource
• GPU Optimization savings estimate in overview
• Fix saved reports Uh Oh! Something went wrong errors
• Fix consistency when sharing namespace overhead costs
• Fix Discrepancy in estimated savings
• Fix GPU feature gate
• Fix GPU utilization showing when no GPU exists
• Fix DataDog costs not displaying properly
• Fix container right-sizing parameters for table/csv download
• Bump cluster controller v0.16.11 (cherry-pick #3795)
• Bump modeling 0.1.19 fix cve-2024-12254 (cherry-pick #3775)

v2.5.2-rc.3

What's Changed

• fix cloud cost filter consistency
• fix container request-sizing pod name display
• fix container request-sizing csv download to export all rows
• fix GPU savings showing idle load balancer as a GPU resource
• GPU Optimization savings estimate in overview
• bump modeling 0.1.19 fix cve-2024-12254 (cherry-pick #3775) by @gcp-cherry-pick-bot in #3798

v2.5.2-rc.2

What's Changed

• bump cluster controller (cherry-pick #3795) by @gcp-cherry-pick-bot in #3796

v2.5.2-rc.1

What's Changed

• Fix saved reports Uh Oh! Something went wrong errors
• Fix consistency when sharing namespace overhead costs
• Fix Discrepancy in estimated savings
• Fix GPU feature gate
• Fix GPU utilization showing when no GPU exists
• Fix DataDog costs not displaying properly
• Fix cotainer right-sizing parameters for table/csv download

v2.5.1

Fixes

• Fix issues running aggregator container in air-gapped/no-internet systems
• Fix issues with load-balancer costs being incorrect
• Add kubecost version to the diagnostics endpoint data
• UI: prevent quantile request-sizing with hourly data
• UI: add consistent units to request-sizing csv
• Change default window for all profiles to 3d for cluster right-sizing

Helm Changes

• bump to kiwigrid 1.28.1 (cherry-pick #3764) by @gcp-cherry-pick-bot in #3767
• add init container for updating ca trust (#3763) by @mittal-ishaan in #3770

Full Changelog: v2.5.0...v2.5.1

v2.5.1-rc.2

• Fix aggregator image revision 3