fix(tests): isolate shared remote-ledger homes between bearings snapshot cases - #170
Merged
prajwal-395 merged 1 commit intoSep 22, 2026
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The remote-ledger fixture homes live under the suite-wide TMP_ROOT, so every case shared the same victim paths. The concurrent-budget case leaves a slow-ledger-read marker on ledger-1 behind, and the next case to build a fleet on that path (e.g. the kindless v1 summary case) read stale instead of fresh and failed - an order-dependent bleed that the committed suite order only masked because an intermediate case happened to clean the marker.
make_remote_ledger_fleet now resets each shared remote home (rm -rf + recreate) so a previous case's control files - slow/unbounded read markers, stale summaries - can never reach another case's snapshot, regardless of execution order.
Proven with test_remote_ledger_fleet_resets_shared_homes_between_cases, which plants the exact leftover and asserts the next case's snapshot stays fresh with its own artifacts: it fails on the pre-fix helper and passes with the fix. Baseline refreshed to count=61 via bin/fm-stock-bash-baseline.sh; fm-lint.sh clean.