Purge-k8s-jobs-cli is a cobra based CLI wrapper over the purgek8sjobs package.
This utility provides a command line tool based interface to the purgek8sjobs, which would be the most appropriate way to consume the package.
The reasons to why anyone might want use this tool are mentioned here
The docker image with the tool's binary is available here
A helm chart is available to enable easy deployment to k8s clusters.
helm upgrade --install purge-k8s-jobs helm
The binary is available as a release artifact here
To build a binary from source:
make build # For OSX
make build-linux # For Linux
make build-linux BINARYNAME=xyz # For custom binary names
For command help text:
./bin/purge-k8s-jobs-cli-v0.1.0 -h
This utility helps purge all kubernetes jobs that have completed their execution more than `b` hours ago in the namespace `n`. Optionally, this also write the job spec and events to a provided log file path
Usage:
purge-completed-k8s-jobs --namespace xyz --before-hours 5 --events-log-file=./events.txt --spec-log-file=./spec.txt
Flags:
-b, --before-hours int16 Query and delete jobs that were complete before x hours (default 1)
-e, --events-log-file string Log file to write the job events to
-h, --help help for purge-completed-k8s-jobs
-n, --namespace string Namespace in which the operations are to be performed (default "default")
-s, --spec-log-file string Log file to write the job spec to
To use kubeconfig file for authentication:
export KUBECONFIG=~/.kube/config
purge-completed-k8s-jobs-v0.1.0 --namespace xyz --before-hours 0 \
--events-log-file=./events.txt --spec-log-file=./spec.txt
To set custom log levels: (The default log level is debug )
export LOG_LEVEL=warn
purge-completed-k8s-jobs-v0.1.0 --namespace xyz --before-hours 0 \
--events-log-file=./events.txt --spec-log-file=./spec.txt