Failure on clustered ducktape of `TestMirrorMakerService`.`test_consumer_group_mirroring`

[jcsp]

This looks like a fairly consistent failure two nights in a row:

FAIL test: TestMirrorMakerService.test_consumer_group_mirroring.source_type=kafka (2/2 runs) failure at 2022-08-01T10:40:48.978Z: TimeoutError('') in job https://buildkite.com/redpanda/vtools/builds/3077#01825803-32b3-4fee-8e12-243a0177833c failure at 2022-07-31T11:04:20.402Z: TimeoutError('') in job https://buildkite.com/redpanda/vtools/builds/3073#018252da-dd82-4dac-bad9-ea37369ded6e FAIL test: TestMirrorMakerService.test_consumer_group_mirroring.source_type=redpanda (2/2 runs) failure at 2022-08-01T10:40:48.978Z: TimeoutError('') in job https://buildkite.com/redpanda/vtools/builds/3077#01825803-32b3-4fee-8e12-243a0177833c failure at 2022-07-31T11:04:20.402Z: TimeoutError('') in job https://buildkite.com/redpanda/vtools/builds/3073#018252da-dd82-4dac-bad9-ea37369ded6e

Since it is not failing this way in docker, likely to be something about the environment.

[jcsp]

Still failing https://buildkite.com/redpanda/vtools/builds/3104#018266cf-8592-4261-83db-4645611ed820 https://buildkite.com/redpanda/vtools/builds/3104#018266cf-8592-4261-83db-4645611ed820

I took a superficial look, the test is timeout out on its main check for sync, there is an 8GB mirror maker log in there that probably needs downloading to find out why (or re-run the test interactively on clustered ducktape and inspect it in-place)

[dlex]

Triage bottomline: failure due to an issue in the test.

Consumer groups described in the source and in the target cluster are equal with the exception of the order of the partitions. This code should be changed to an order independent comparison.