Tool for dumping manifests from your Kubernetes clusters.
Binaries are available for all major platforms. See the releases page.
Using the Homebrew package manager for macOS:
brew install sj14/tap/kubedump
It's also possible to install via go install:
go install github.com/sj14/kubedump@latest
See deploy/cronjob.yaml as an example how to deploy a CronJob with kubedump. You have to adjust the file accordingly, for example to push the dumped data to a persistent storage.
Usage of kubedump:
-clusterscoped
dump cluster-wide resources (default true)
-config string
path to the kubeconfig, empty for in-cluster config (default "~/.kube/config")
-context string
context from the kubeconfig, empty for default
-dir string
output directory for the dumps (default "dump")
-groups string
groups to dump (e.g. 'metrics.k8s.io,coordination.k8s.io'), empty for all
-ignore-groups string
groups to ignore (e.g. 'metrics.k8s.io,coordination.k8s.io')
-ignore-labels string
ignore resources with the given labels (e.g. key1=value1,key2=value2)
-ignore-namespaces string
namespaces to ignore (e.g. 'ns1,ns2')
-ignore-resources string
resources to ignore (e.g. 'configmaps,secrets')
-labels string
dump resources with the given labels (e.g. key1=value1,key2=value2), empty for all
-namespaced
dump namespaced resources (default true)
-namespaces string
namespaces to dump (e.g. 'ns1,ns2'), empty for all
-resources string
resources to dump (e.g. 'configmaps,secrets'), empty for all
-stateless
remove fields containing a state of the resource (default true)
-threads uint
maximum number of threads (minimum 1) (default 10)
-verbosity uint
verbosity of the output (0-3) (default 1)
-version
print version information of this release
All options can also be set as environment variables by using their uppercase flag names and changing dashes (-) with underscores (_), e.g. ignore-namespaces becomes IGNORE_NAMESPACES.