Skip to content

fix: batch catalog scans and defer status aggregates - #178

Merged
Yehonal merged 1 commit into
mainfrom
pr/amf-catalog-batch
Oct 26, 2026
Merged

Yehonal merged 1 commit into
mainfrom
pr/amf-catalog-batch

Conversation

@Yehonal

@Yehonal Yehonal commented Oct 26, 2026

Copy link
Copy Markdown
Contributor

Catalog scans previously committed every queue operation and computed full inventory aggregates on every service tick. Each bounded scan now commits metadata, text, coverage and queue cursors together, while background ticks use constant-size scan state and minimal scan results. Default public scan responses retain their existing status fields. Directory traversal inspects newly discovered children before continuing enumeration within a batch.

Rollback closes advanced directory iterators so retries resume from committed cursors. No schema, policy, or SQLite durability settings change.

Validation: focused catalog/admin tests passed 21/21; full npm test passed 1,207 tests with 9 skips and zero failures (1,216 total). Regressions cover one transaction per batch, cursor/iterator rollback, atomic metadata/FTS/state rollback, SIGKILL during an open transaction, private/no-follow protections, and background ticks without eager aggregates. git diff --check and Syncwheel validation passed. Independent review reported no findings.

Rollout: the administrative service needs the updated source and a restart after merge. This PR does not change running services.

@Yehonal
Yehonal merged commit f20301f into main Oct 26, 2026
1 check passed
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