John Spray

CI failures: - https://github.com/redpanda-data/redpanda/issues/5378 - https://github.com/redpanda-data/redpanda/issues/5836 - https://github.com/redpanda-data/redpanda/issues/5837

Thanks for making it through a lengthy set of commits. All the comments I've silently marked resolved are addressed in https://github.com/redpanda-data/redpanda/pull/5970

CI failures were a transient issue apparently https://redpandadata.slack.com/archives/C02LZGSS66M/p1659967842254759

CI failure is: - https://github.com/redpanda-data/redpanda/issues/5950

/backport v22.2.x

This needs a clustered ducktape run before merging to see which existing tests might be tripping bad_allocs routinely

> @jcsp - seems reasonable, is that something I can get CI to do for me? Maybe, ask #help-devprod to write some instructions into the wiki if they didn't already....

This LGTM once we know whether it immediately fails on any of the nightly tests in scale_tests (also needs a rebase).

(confession: I'm partly motivated by wanting to just do fewer retries from my test code that uses rpk on large systems, but I think this is a genuine UX issue...

I'll try... currently this is happening in the middle of big automated runs so it'll be a cause of pausing something on a large system that's in a wonky state...