In order to better understand and attribute cost of my Kubernetes cluster, I am looking for a script written in Perl (preferrable) or Golang that will run and query the cluster to report back on CPU, RAM, and DISK usage and allocation to all of the different pods, namespaces, etc, and breakdown the usage by namespace, pod, and any number and combination of labels.
This script can be scheduled to run at regular intervals to collect the metric data over time. This data (csv formatted output file) will be consumed by a separate process (not part of this project) to generate daily, weekly, monthly, etc usage reports.
I should be able to supply input parameters to run the script to target the entire cluster or a specific namespace, for example. And I should be able to also provide the format of the output such that I can specify the usage allocation broken down by a specific label (i.e. label=Team) or a combination of labels (i.e. label=Team and label=Environment).
It will then generate a csv file similar to:
Team,Environment,CPU,Memory,Disk,Total CPU, Total Memory, Total Disk
"CPU,Memory,Disk" indicate resources used by Team & Environment label and
"Total CPU, Total Memory, Total Disk" indicate total cluster resources
I also welcome your feedback and suggestion for improvement and clarification.
8 freelancers are bidding on average $594 for this job
Hello! if you want collect metrics from your cluster, you should use kibana and prometheus for it, dont use perl script - this is old solutions , it's not modern.
Can help You create such script, which collect/save your Kubernetes cluster stats into external cvs file (which can be run by crontab). Tnx and Regards.
Hello, we would like to offer you cooperation on the project, according to your requirements. Please, contact us for more information. You can also find us at www.Pexxi.eu. Thanks and look forward to our cooperation.