Kuberhealthy Versions Save

A Kubernetes operator for running synthetic checks as pods. Works great with Prometheus!

v2.8.0-rc2

4 months ago

What's Changed

Full Changelog: https://github.com/kuberhealthy/kuberhealthy/compare/v2.8.0-rc1...v2.8.0-rc2

v2.8.0-rc1

4 months ago

Test release to test the new deployment github workflow

v2.7.1

2 years ago

This release mostly contains minor fixes and documentation. Special thanks to @Hungrylion2019 this release. Thank you for all your help!

If your business uses Kuberhealthy, please take a moment to add it to the adopters list! This helps us advance through the ranks of the CNCF.

What's Changed

New Contributors

Full Changelog: https://github.com/kuberhealthy/kuberhealthy/compare/v2.7.0...v2.7.1

v2.7.0

2 years ago

Version v2.7.0 raises the minimum supported version of Kubernetes to 1.16.

  • Fixes for checks failing to create khstate resources #988. (@jonnydawg)

v2.6.0

2 years ago

For this v2.6.0 release, apply the newly generated CRDs:

  • KuberhealthyCheck
  • KuberhealthyJob
  • KuberhealthyState

Bug Fixes:

  • Fixing prometheus rules integration for operator and additionalLabels #962 @skuethe
  • Remove Comcast mentions #969 @integrii
  • Reaper NPE bug - ensure container terminated state exists before fetching container terminated time #970 @joshulyne
  • Set owner references only on pods in the kuberhealthy namespace #903, PR #973 @mikeinton
  • Add missing namespace metadata to helm chart configmap #976, PR #977 @integrii

New Features:

  • Generate and update kuberhealthy crds #778, PR #937 @joshulyne
    • Include node that checks ran on in khstate data #351

v2.5.0

2 years ago

Kuberhealthy v2.5.0

Bug Fixes:

  • khcheck-storage fails due to template issues - #895, PR #896 @blame19
  • Kuberhealthy does not work properly on minikube - #866, PR #884 @integrii, @joshulyne
  • Cannot unmarshal string into Go struct field Config.JobCleanupDuration of type time.Duration - #857, PR #884 @integrii, @joshulyne
  • Refetch khstate CRD when storing the khstate for a check #936 @joshulyne
  • Ensure only master kuberhealthy pod restarts khchecks when config is reloaded #942 @joshulyne
  • Fix URL for latest helm release #952 @dmaljovec
  • Fix restartPolicy indent on a few khcheck yaml specs #948 @joshulyne
  • Fix for checkReaper and configReloader bug #955 @joshulyne

New Features:

  • Kuberhealthy has been accepted into the CNCF Sandbox 🎉
  • Migration from Comcast github organization to its own (https://github.com/kuberhealthy) 🎉
  • Added optional support for Kuberhealthy pod affinity/antiaffinity - PR #909, @AshutoshNirkhe
  • Added namespace to pod status error messages - PR #902, @rawlingsj
  • Added option to install pod restart checks at the cluster scope - PR #897, @rawlingsj
  • Refactored external report validation to use check run uuid, but maintain support for also using pod IP validation PR #884, @joshulyne, @integrii
  • Refactor config options for checkReaper: maxKHJobAge, maxCheckPodAge, maxCompletedPodCount, maxErrorPodCount and allow for 0 pod count values or the option to have no completed pods lying around - PR #884, @joshulyne, @integrii
  • Up minor versions on all checks to use new checkClient package #944 @mikeinton

v2.4.1

3 years ago

Kuberhealthy v2.4.1

Bug Fixes:

  • Fix for retries when updating the kuberhealthy run UUID on the khstate resource - #858, PR #872 @integrii
  • Fix checkReaper duration configs, add failedPodCleanupDuration making it configurable, re-enable unit tests for kuberhealthy - #857, #859, #861, PR #862 @joshulyne
  • Edit prometheus operator installation instructions; fix grafana prometheus datasource - #848, #850 @joshulyne
  • Set proper pod port where prometheus metrics are exposed - PR #869 @randrusiak
  • Pod-restarts check - fixed updating errorMessages with err, fixed pod-restarts-check rbac role to be able to list events - #828, PR #829 @isaaguilar

New Features:

  • Complete refactor of ssl_util package for ssl-hanshake check - #879 @integrii

New Kuberhealthy Check:

v2.4.0

3 years ago

Kuberhealthy 2.4.0

Upgrade Instructions:

When upgrading to this release, you should expect to see the check-reaper cronjob be removed. This functionality is now present in the main Kuberhealthy binary/pod. If you deployed via helm template, you must manually remove the check-reaper cronjob and its associated RBAC. This should happen automatically if you used helm install.

Changes and Fixes

  • NPE fix in crd.go #809 @integrii
  • More contexts implemented #640 @integrii
  • Prometheus metrics compatibility with DataDog #813 @joshulyne
  • NPE fix in deployment check #796 @HeroBcat
  • Move check reaper into main kuberhealthy binary #774 @integrii
  • Fixes in deployment check tolerations #783 @jonnydawg @jdowni000
  • All checks rebuilt and pushed with latest versions #820 @joshulyne
  • Removed tolerations for cordoned nodes from the daemonset check #785 @joshulyne

v2.3.2

3 years ago

Kuberhealthy 2.3.2

Upgrade Instructions:

  • When upgrading to this release, you must be sure the khjob custom resource is applied in your cluster. Without this, the check reaper will crash. If you use the Helm chart, this will be done for you automatically.

Features:

  • Updated labels for daemonset checks to be less generic. #685, #723 @zjhans
  • Helm ServiceMonitor endpoint specifications parameterized. #726, #731 @geoL86
  • Added env variables and a new function to allow scheduling of checker pods on dedicated nodes. #722, @jdowni000
  • Added toleration support for checks. #728, #741 @zjhans
  • Added registry override options at global level and for individual checks. #734, #751 @zjhans
  • Added ServiceMonitor custom labels. #742, #757 @zjhans
  • Added override options for check reaper thresholds. #768, #769 @AshutoshNirkhe
  • Enhanced http-check to support other request types besides GET, and add configurable expected status code. #779, #782 @yashvardhan-kukreja

Bug Fixes:

  • Added ACTIONS_ALLOW_UNSECURE_COMMANDS env var to fix deprecated GitHub 'add path' action. #727, #729 @jonnydawg
  • Added check-reaper ClusterRole to fix permissions issue. #732, #737 @joshulyne
  • Updated invalid, legacy URL in Helm chart. #736, #738 @TJM
  • Removed nodecheck WaitForKubeProxy for clusters not using kube-proxy, and modified WaitForNodeAge to pass in nodeName using downward API instead of needing additional permissions to get nodeName information. #739, #752, #755 @joshulyne
  • Set default PASSING_PERCENT for http-check. #768, #780 @jdowni000

v2.3.1

3 years ago

Kuberhealthy 2.3.1

Bug Fixes:

  • Manually set KH_EXTERNAL_REPORTING_URL environment variable is now properly processed and respected #715 @zjhans
  • ConfigMap with Prometheus rules had been deprecated, updated to be properly picked up #718 @linkvt
  • The whitespace between both ConfigMaps was chomped, causing the ConfigMaps to merge, this has been fixed. #718 @linkvt