Cache the STDOUT of a given command, for no more than a given duration (default is 5 minutes).

NOTE: I created this with kubectl get pod in mind. It's not specific to kubectl (part of the kubernetes project) but it would need some more work to become more generally useful.

NOTE: This is totally experimental. Only use it for something with smallish output, and be careful with write commands.

go get -u github.com/laher/clicache

Call this command twice, and the second time will be super quick:

clicache kubectl --context=dev get pod
clicache kubectl --context=dev get pod

The initial call takes about 30 seconds, the second invocation takes milliseconds.

You can also delete entries with -del:

clicache -del kubectl --context=dev get pod

For more options, use clicache -h

I include the following functions in my zsh config (hopefully it works in bash too)

function ki {
  clicache kubectl --context=$1 get pod | grep $2 | cut -d ' ' -f1
}

function k {
  kubectl --context=$1 ${@:2}
}

function kclc {
  clicache -del kubectl --context=$1 get pod
}

Now, given a context dev and a pod named mypod I can quickly run commands like this (without clicache there would be a few seconds delay fetching the container name):

k dev log -f $(ki dev mypod)
k dev exec -it $(ki dev mypod) sh

In the future I'd like to wrap kclc dev into some other commands, in order to explicitly clear cache