CronJob Cleaner

Why

I wrote this cleaner as a way to delete pods for jobs that got "stuck", jobs that just seemed to stop doing anything but the job and pod were both in a running state for hours. The activeDeadlineSecond would delete the job, but this wouldnt resolve any alerts from alert manager. This project is my fix, and a good excuse to write go.

How

Environment Variables

Environment Variable	Type	Description
InCusterConfig	bool	Used to determine which clientset should be used for kubernetes authentication. Defaults to false, set to true to use inside a cluster as a cronjob.
ActiveDeadlineSecond	float64	Used to determine which jobs/pods should be identified for cleanup. Defaults to 4200.
GetPodLogs	bool	Used to determine if the pods logs should be relogged as Information. Defaults to false.
LogTail	int64	Used to determine the tail of logs that should be relogged, if GetPodLogs is set to true, otherwise it is ignored. Defaults to 100.
CleanFailedJob	bool	Used to determine if jobs that are in a failed state should be identified for cleanup. DeleteJob must be true for these jobs to be deleted. Defaults to false.
DeleteJob	bool	Used to determine if jobs that are identified for cleanup should be deleted. Defaults to false.
DeletePod	bool	Used to determine pods that are identified for cleanup should be deleted. Defaults to false.
ExcludedNamespaces	string	Comma separted list of namespaces to NOT clean.

Run as cronjob in kubernetes

Use Kustomize

curl -s "https://raw.githubusercontent.com/\
kubernetes-sigs/kustomize/master/hack/install_kustomize.sh"  | bash

chmod u+x ./kustomize

cp -a ./manifests/. .

./kustomize edit set image kevinshelaga/cronjob-cleaner:tag=kevinshelaga/cronjob-cleaner:latest

./kustomize build . | kubectl -n monitoring apply -f -

Use Helm

helm upgrade cronjob-cleaner ./helm -n monitoring --create-namespace --install

Whats left

TODO

• More/better tests
• Helm chart
• Policies around contributions