Ketall is too slow when there are many objects in the cluster

[corneliusweig]

No description provided.

[tillig]

Just hit this myself. I had previously written a script that basically did:

• kubectl api-resources -o name --namespaced=true (or --namespaced=false, depending on what I'm looking for)
• Removes anything that looks like events
• Joins everything in a comma-delimited list
• Does kubectl get list,of,comma,delimited,resources

...and it appears to be a lot faster than Ketall. Ketall takes 1m 49s to list the resources in a namespace; same namespace with the shell script (same resources displayed) is 16s. Adding --use-cache doesn't seem to make more than a couple seconds difference.

Granted, Ketall is doing more with respect to filtering, etc., but I'm not sure if maybe that sparks some ideas or something.

Here's my script in case you're interested.