fix(cloudflare): use timestamp-based ids to handle table recreation

[dinwwwh]

Summary

Fixes an issue where resuming from a message id after idle cleanup or schema reset could skip newer events. The problem occurs because SQLite's AUTOINCREMENT sequence restarts when a table is recreated, causing new ids to be lower than ids issued before the cleanup.

Changes

• Use timestamp-based IDs: Initialize the sqlite_sequence to Date.now() * 1000 (microseconds) when creating the events table, ensuring new ids always exceed previously issued ids
• Update tests to handle dynamic IDs:
  • Changed hardcoded id assertions (e.g., '1', '2') to use expect.any(String) or compute relative to the first received id
  • Added test case resumes events stored after idle cleanup for an id issued before it to verify the fix works correctly
  • Added setup code in resumes messages in numeric id order test to reset the sequence and test the digit boundary scenario

Implementation Details

The fix ensures that when a durable object's storage is cleaned up and the events table is recreated, the AUTOINCREMENT sequence starts at a microsecond timestamp rather than 1. This guarantees that any new events published after cleanup will have ids greater than ids issued before cleanup, making resume queries (WHERE id > ?) work correctly.

https://claude.ai/code/session_017T4x44m5ATzLkRn6JNgLr2

[cloudflare-workers-and-pages]

Deploying with    Cloudflare Workers

The latest updates on your project. Learn more about integrating Git with Workers.

Status	Name	Latest Commit	Preview URL	Updated (UTC)
✅ Deployment successful! View logs	orpc	94c99db	Commit Preview URL Branch Preview URL	Oct 06 2026, 08:18 AM

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[codspeed]

Merging this PR will not alter performance

✅ 28 untouched benchmarks
⏩ 2 skipped benchmarks¹

---

_{Comparing claude/brave-newton-hkj1wr (94c99db) with main (769916d)}

Footnotes

1. 2 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[pkg-pr-new]

More templates

• @orpc/bun-playground
• @orpc/cloudflare-playground
• @orpc/nest-playground
• @orpc/next-playground

@orpc/ai-sdk

npm i https://pkg.pr.new/@orpc/ai-sdk@2190

@orpc/arktype

npm i https://pkg.pr.new/@orpc/arktype@2190

@orpc/bun

npm i https://pkg.pr.new/@orpc/bun@2190

@orpc/client

npm i https://pkg.pr.new/@orpc/client@2190

@orpc/cloudflare

npm i https://pkg.pr.new/@orpc/cloudflare@2190

@orpc/contract

npm i https://pkg.pr.new/@orpc/contract@2190

@orpc/experimental-effect

npm i https://pkg.pr.new/@orpc/experimental-effect@2190

@orpc/evlog

npm i https://pkg.pr.new/@orpc/evlog@2190

@orpc/hibernation

npm i https://pkg.pr.new/@orpc/hibernation@2190

@orpc/json-schema

npm i https://pkg.pr.new/@orpc/json-schema@2190

@orpc/experimental-lock

npm i https://pkg.pr.new/@orpc/experimental-lock@2190

@orpc/experimental-msw

npm i https://pkg.pr.new/@orpc/experimental-msw@2190

@orpc/nest

npm i https://pkg.pr.new/@orpc/nest@2190

@orpc/next

npm i https://pkg.pr.new/@orpc/next@2190

@orpc/node

npm i https://pkg.pr.new/@orpc/node@2190

@orpc/openapi

npm i https://pkg.pr.new/@orpc/openapi@2190

@orpc/opentelemetry

npm i https://pkg.pr.new/@orpc/opentelemetry@2190

@orpc/pinia-colada

npm i https://pkg.pr.new/@orpc/pinia-colada@2190

@orpc/pino

npm i https://pkg.pr.new/@orpc/pino@2190

@orpc/publisher

npm i https://pkg.pr.new/@orpc/publisher@2190

@orpc/ratelimit

npm i https://pkg.pr.new/@orpc/ratelimit@2190

@orpc/server

npm i https://pkg.pr.new/@orpc/server@2190

@orpc/shared

npm i https://pkg.pr.new/@orpc/shared@2190

@orpc/swr

npm i https://pkg.pr.new/@orpc/swr@2190

@orpc/tanstack-query

npm i https://pkg.pr.new/@orpc/tanstack-query@2190

@orpc/trpc

npm i https://pkg.pr.new/@orpc/trpc@2190

@orpc/valibot

npm i https://pkg.pr.new/@orpc/valibot@2190

@orpc/zod

npm i https://pkg.pr.new/@orpc/zod@2190

commit: 94c99db

[pullfrog]

✅ No new issues found.

Reviewed changes — reviewed the full PR head (all four commits, final state), including a local test run of both publisher suites and type:check.

• Time-seeded AUTOINCREMENT sequence — when the events table is actually created (rowsWritten > 0), ensureSchemaAndCleanup now seeds sqlite_sequence with Date.now() * 1000 microseconds, so ids issued after recreation are always greater than ids issued before it. Plain integers are preserved, so the existing WHERE id > ? numeric comparison and the TEXT-cast read path are untouched.
• Regression test — resumes events stored after idle cleanup for an id issued before it fails without the seeding (replays 0 instead of 2) and passes with it, so it genuinely pins the reported bug.
• Test adaptations — hardcoded ids ('1', '2', '3') became expect.any(String) or computed relative to the first id, and the digit-boundary test now forces UPDATE sqlite_sequence SET seq = 0 to exercise numeric ORDER BY with 9→10.

Verified locally: pnpm --filter @orpc/cloudflare exec vitest run src/publisher-object.test.ts src/publisher.test.ts → 34 passed; type:check passes; no stale references to the reverted generation/prefix:meta approach remain.

  ^{｜ View workflow run ｜ Using deepseek-v4.1-flash (free via Pullfrog for OSS) ｜ 𝕏}