Labelling this PR as size/S

@kubernetes/kube-iam @derekwaynecarr @wojtek-t

I would avoid deleteall and favor deletecollection since I can imagine DeleteOptions taking a label selector or field selector in the future to scope the set of things it deletes to those that matched.

updated

GCE e2e test build/test passed for commit 777a66f0269200d289671d14d700ba878f36204b.

• Build Log
• Test Artifacts
• Internal Jenkins Results

GCE e2e test build/test passed for commit 2cea0048ac2930ba8963b5a83bd07eb7db833bf7.

• Build Log
• Test Artifacts
• Internal Jenkins Results

Seems like a big change this late for 1.2, but LGTM unless @erictune sees any problems.

GCE e2e test build/test passed for commit 3038eec.

• Build Log
• Test Artifacts
• Internal Jenkins Results

Sorry for the lateness, I didn't notice the addition when it first went in. Delete currently requires knowledge of the resource name, whereas deletecollection does not, which is the main reason it needs a different verb (apart from the get/list congruence)

For context, the namespace controller makes heavy use of delete collection now during namespace termination.

LGTM - thanks @liggitt

lgtm.

GCE e2e test build/test passed for commit 3038eec.

• Build Log
• Test Artifacts
• Internal Jenkins Results

would like this considered for 1.2 before we release a version with collection deletion

Ref #10217

LGTM

@cjcullen FYI there is now a new verb for when you delete several things rather than a named thing.

@k8s-bot test this

Tests are more than 48 hours old. Re-running tests.

This appears safe for ABAC users because the delete verb is not exposed to them, and the IsReadOnly check does not look at delete.

GCE e2e test build/test passed for commit 3038eec.

• Build Log
• Test Artifacts
• Internal Jenkins Results

@k8s-bot test this [submit-queue is verifying that this PR is safe to merge]

GCE e2e test build/test passed for commit 3038eec.

• Build Log
• Test Artifacts
• Internal Jenkins Results

Automatic merge from submit-queue