Skip to content

Finish reorg recovery without per-launch pool request bursts - #57

Merged
droplister merged 1 commit into
mainfrom
fix/recovery-pool-snapshot
Oct 4, 2026
Merged

droplister merged 1 commit into
mainfrom
fix/recovery-pool-snapshot

Conversation

@droplister

Copy link
Copy Markdown
Member

Recovery rechecked every historical refunded launch with a separate pool lookup. Production repeatedly hit Counterparty's rate limit before its final block check, leaving recovery unfinished even after the graduate and graveyard rows were restored.

Read the complete pool listing once during recovery, follow every cursor, and use that snapshot for all closed launches. Failed pages, malformed responses, and repeated cursors leave pool status unknown and preserve existing phases/reserves; ordinary indexing keeps its existing incremental behavior.

Validation: npm run check passed (1,452 tests, 17 render-read checks, type/lint/numeric/i18n checks). Regression covers 208 refunded launches, a failed snapshot followed by successful recovery, preserved graduate reserves, cleared recovery markers, and a single pool-list request. Pagination tests cover later-page pools and incomplete listings.

@droplister
droplister merged commit 4201023 into main Oct 4, 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