lots of Crucible test failures due to timeouts

I'm running the test suite locally for the first time in a while and ran into a log of test Crucible-related test failures that I'd expect to be unrelated to my changes:

$ cargo nextest run --no-fail-fast
...
────────────
     Summary [1935.375s] 2194 tests run: 2183 passed (95 slow), 11 failed, 3 skipped
        FAIL [  53.948s] omicron-nexus::test_all integration_tests::crucible_replacements::test_delete_volume_region_replacement_state_replacement_done
        FAIL [  81.001s] omicron-nexus::test_all integration_tests::crucible_replacements::test_delete_volume_region_replacement_state_requested
        FAIL [  53.394s] omicron-nexus::test_all integration_tests::crucible_replacements::test_delete_volume_region_replacement_state_running_on_pantry
        FAIL [  53.295s] omicron-nexus::test_all integration_tests::crucible_replacements::test_delete_volume_region_snapshot_replacement_state_requested
        FAIL [  52.203s] omicron-nexus::test_all integration_tests::crucible_replacements::test_delete_volume_region_snapshot_replacement_state_running
        FAIL [  51.029s] omicron-nexus::test_all integration_tests::disks::test_disk_expunge
        FAIL [  50.192s] omicron-nexus::test_all integration_tests::disks::test_disk_slot_assignment
        FAIL [  51.146s] omicron-nexus::test_all integration_tests::volume_management::test_delete_image_order_5
        FAIL [  72.444s] omicron-nexus::test_all integration_tests::volume_management::test_read_only_region_reference_counting
        FAIL [  47.413s] omicron-nexus::test_all integration_tests::volume_management::test_read_only_region_reference_usage_sanity
        FAIL [  11.927s] omicron-omdb::test_all_output test_omdb_usage_errors
error: test run failed

(the omdb one is definitely mine)

This is all on commit ad09008d11a73817571581ce179bc96e1c480fa8 (which is far from "main", just not in ways I'd expect to affect this). I will double check whether this happens in CI and whether I see it on "main" when running locally too.

Start by reproducing the failures with cargo nextest run --no-fail-fast at commit ad09008d11a73817571581ce179bc96e1c480fa8. Inspect the listed integration_tests::crucible_replacements, disks, and volume_management tests separately from test_omdb_usage_errors; done means the Crucible-related timeout failures are explained or fixed and the relevant tests pass.