Skip to content

integration-test: latest test are processed with batch block #579

Open
lupin012 wants to merge 7 commits into
mainfrom
lupin012/latest_test_in_batch
Open

integration-test: latest test are processed with batch block #579
lupin012 wants to merge 7 commits into
mainfrom
lupin012/latest_test_in_batch

Conversation

@lupin012

Copy link
Copy Markdown
Collaborator

When running -L with a live reference node (-e/-L), tests are now dispatched in batches of N (default 50) with a sync check before each batch. N=0 keeps the previous behaviour (single sync then all at once).

lupin012 and others added 7 commits June 18, 2026 14:44
When running -L with a live reference node (-e/-d), tests are now
dispatched in batches of N (default 50) with a sync check before each
batch. N=0 keeps the previous behaviour (single sync then all at once).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
If a batch has failures and the run is not already aborting, reset the
batch index to 0 so all tests re-run from a fresh sync. The natural
stop conditions (-M max failures, ExitOnFail) still apply.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Makes it immediately visible whether a batch belongs to the first run
or a retry triggered by a previous failure.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Check ctx cancellation and max-failures at the start of each iteration
rather than burying them inside the restart condition, making control
flow easier to follow.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Give the testbed node more time to catch up with the reference node
before declaring sync failure.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
When a batch fails and we restart from attempt 1, restore stats and
report entries to their pre-attempt snapshot so the final summary
reflects only the last complete attempt, not accumulated counts from
previous failed attempts.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
When a latest-block batch has test failures, stop immediately and exit
with code 1 instead of restarting internally. The shell retry loop
handles re-running from batch 1, which keeps the retry logic in one
place and avoids the potential infinite-loop of unbounded internal
attempts.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant