88 votesRob Archibald supported this idea ·Rob Archibald commented
Agreed. Didn't see this when I posted mine at https://googlecloudplatform.uservoice.com/forums/302601-container-engine/suggestions/33052594-make-it-so-that-gcepersistentdisk-or-another-persi. It seems ridiculous that we have to roll our own ReadWriteMany storage.
51 votesRob Archibald shared this idea ·
4 votesRob Archibald commented
You can simply set the replicas to 0 in the yaml and it'll effectively turn off the cluster. And, it'd then be a fairly simple matter of scaling it back up the same way. So, if you wanted to, you could set up a job to run "kubectl apply -f your-deployment-day.yaml" during the day to start the cluster and kubectl apply -f your-development-night.yaml" to stop it during evening and weekends.