use new 'kubectl get all' ?

[joemiller]

kubectl get now has an 'all' flag.

[spheromak]

I think so prob should PR that, cause as-is it is going to miss stuff

[joemiller]

a couple issues:

1. it doesn't actually 'get all' resources, the name is a lie. We know this from our styx and valhalla makefiles 'status-cluster' tasks, although I forget which resources it wasn't pulling. This might be fixed in kubectl 1.6 though. It appears that it still doesn't get TPRs though.

2. get all means we would lose the ability to split up individually by object. We know from our only recovery event that being able to individually recover objects was critical. If they're all lumped into a giant file then you need to split them up. We can do that at backup time perhaps by adding additional code. I don't want to have to do that at recovery time though when time is a critical, scarce resource.