List Kubernetes PODs of a specific namespace or all namespaces with their configured memory or cpu requests, limits and the current memory or cpu usage.
- Overview of PODs and their memory or cpu requests, limits and actual usage
- Provides an easy inspection of the cluster utilization in terms of memory or cpu
- Uses
kubectlunder the hood and reuses its config - Supports
--namespaceand--all-namespacescommand line arguments - Supports
--contextcommand line argument filters and column setup
Example:
[15:09:58] enrico@host (0): ~% kubecargoload.py --all-namespaces
Namespace Name Requests Limits Usage %
----------------------------------------------------------------------------------------------------------
default kube-web-view-7c67ddb647-pvjvs 100.0 Mi 100.0 Mi 29.0 Mi 29.00 %
jitsi jitsi-57d5888c88-vzrzl 0.0 B 0.0 B 137.0 Mi 0.00 %
kube-system coredns-66bff467f8-qn4pq 70.0 Mi 170.0 Mi 13.0 Mi 7.65 %
kube-system coredns-66bff467f8-znpxv 70.0 Mi 170.0 Mi 11.0 Mi 6.47 %
kube-system etcd-minikube 0.0 B 0.0 B 58.0 Mi 0.00 %
kube-system kindnet-ptgnz 50.0 Mi 50.0 Mi 13.0 Mi 26.00 %
kube-system kube-apiserver-minikube 0.0 B 0.0 B 257.0 Mi 0.00 %
kube-system kube-controller-manager-minikube 0.0 B 0.0 B 50.0 Mi 0.00 %
kube-system kube-proxy-s7bl8 0.0 B 0.0 B 16.0 Mi 0.00 %
kube-system kube-scheduler-minikube 0.0 B 0.0 B 19.0 Mi 0.00 %
kube-system metrics-server-7bc6d75975-d6sgr 0.0 B 0.0 B 13.0 Mi 0.00 %
kube-system nginx-ingress-controller-6d57c87cb9-tgwwm 0.0 B 0.0 B 0.0 B 0.00 %
kube-system storage-provisioner 0.0 B 0.0 B 15.0 Mi 0.00 %
----------------------------------------------------------------------------------------------------------
Summary (PODs without configured limits ignored) 290.0 Mi 490.0 Mi 66.0 Mi 13.47 %
In order to use kube-cargo-load, you will need:
- A Kubernetes cluster to connect to
- Kubernetes Metrics Server must be installed and running in Kubernetes (https://github.com/kubernetes-sigs/metrics-server)
- kubectl (it must be configured for your Kubernetes cluster)
- Python 3.6 or newer
The easiest method is to install directly from pypi using pip:
pip install kube-cargo-load
If you prefer, you can download kube-cargo-load and install it directly from source:
python setup.py install
Alternatively, you can download just the script and execute it:
wget https://raw.githubusercontent.com/eht16/kube-cargo-load/master/kubecargoload.py
chmod +x kubecargoload.py
./kubecargoload.py
usage: kubecargoload.py [-h] [-A] [--context CONTEXT] [-d] [-n NAMESPACE] [-H] [-s SORT] [-V]
optional arguments:
-h, --help show this help message and exit
-A, --all-namespaces list the requested object(s) across all namespaces (default: False)
-c, --cpu show cpu instead of memory (default: False)
--context CONTEXT the name of the kubeconfig context to use (default: None)
-d, --debug enable tracebacks (default: False)
-n NAMESPACE, --namespace NAMESPACE
namespace to use (default: default)
-H, --no-headers do not print header line before the output (default: False)
-s SORT, --sort SORT sort by column(s), to sort by multiple columns seperate them with comma. Valid options: namespace,name,requests,limits,usage,ratio (default: namespace,name)
-V, --version show version and exit (default: False)
The source code is available at https://github.com/eht16/kube-cargo-load/.
* Add support for -c/--cpu flag (#3, Kurounin)
* Round humanized usage values to have more accurate numbers
* Add unit tests
* Fix Python 3.6 compability (#1, Adrian Gajan)
* Improve docs
* Fix display units (KB -> Ki)
* Consider only PODs with configured limits in the summary
* Initial release
Found a bug or got a feature request? Please report it at https://github.com/eht16/kube-cargo-load/issues.
Enrico Tröger enrico.troeger@uvena.de