Skip to content
TheChatArchiveThe Chat Archive
Claudeclaude-sonnet-5· anthropicHugging Face ibragim-bad/swe_rebench_07_2026_trajectories

ISSUE DESCRIPTION magic-link verify deadlocks: runWithTransaction nests…

  1. User

    ISSUE DESCRIPTION

    magic-link verify deadlocks: runWithTransaction nests adapter.transaction (1.6.11 consumeVerificationValue) instead of reusing the active transaction

    Is this suited for github?

    • Yes, this is suited for github

    Reproduction

    1. better-auth 1.6.11 with the magicLink plugin.
    2. A database adapter whose transaction is a real interactive DB transaction that holds a dedicated connection for its duration (e.g. the Kysely adapter, or @zenstackhq/better-auth, backed by pg).
    3. Limit the client connection pool to one connection: new Pool({ max: 1 }) -- standard when using Cloudflare Hyperdrive, which pools server-side so the worker keeps a tiny client pool.
    4. await auth.api.signInMagicLink({ body: { email } }) -> works (single insert).
    5. await auth.api.magicLinkVerify({ query: { token } }) (or GET /api/auth/magic-link/verify?token=...).

    Minimal trigger: any setup where adapter.transaction acquires a connection and the pool cannot hand out two connections at once.

    Current vs. Expected behavior

    Current: the verify call hangs forever and never returns. Postgres shows the backend stuck idle in transaction (wait_event = ClientRead) right after the verification SELECT; the connection is never released, so every attempt leaks one idle-in-transaction connection.

    Trace of the cause:

    • internalAdapter.consumeVerificationValue (added in 1.6.11) runs inside runWithTransaction -> opens transaction tx1 (BEGIN, takes the only pooled connection).
    • Consuming the row (consumeOne) triggers another runWithTransaction -> calls adapter.transaction again -> transaction tx2, which tries to acquire a second connection.
    • runWithTransaction in @better-auth/core (context/transaction.mjs) always opens a fresh adapter.transaction; it does NOT check adapterAsyncStorage.getStore() to see whether a transaction is already active in the current async context. So tx2 is nested inside tx1 and blocks forever waiting for the connection tx1 holds -> deadlock.

    Note the CRUD path already reuses the active transaction: getCurrentAdapter(adapter) returns the trx adapter from the ALS store. runWithTransaction is inconsistent with this -- it doesn't reuse.

    Expected: runWithTransaction should be reentrant -- when an adapter/transaction is already active in the async-context store, run fn within it instead of opening a nested adapter.transaction (mirroring getCurrentAdapter). Then verify needs only one connection and works on any pool size. (Alternatively/additionally, document that adapters MUST support reentrant nested transactions -- currently adapters like @zenstackhq/better-auth open a brand-new transaction on the base client for each adapter.transaction call, which is a reasonable contract but incompatible with the current nesting.)

    This is security-relevant: the 1.6.11 change that introduced this nesting was the magic-link "mint multiple sessions from one token" race fix, so the fix effectively breaks magic-link verification on connection-limited adapters/pools.

    What version of Better Auth are you using?

    1.6.11

    System info

    > The `auth info` CLI can't introspect this project headlessly because the auth instance is built dynamically via a factory (no static `auth.ts` export), so the values below are assembled manually
    
    {
      "better-auth": { "version": "1.6.11", "@better-auth/core": "1.6.11" },
      "adapter": { "name": "@zenstackhq/better-auth", "version": "3.6.4", "orm": "@zenstackhq/orm 3.6.4 (Kysely 0.28.17)" },
      "database": { "driver": "pg 8.20.0", "pool": { "max": 1 }, "fronted_by": "Cloudflare Hyperdrive" },
      "runtime": { "platform": "Cloudflare Workers (workerd)", "wrangler": "4.95.0", "node_for_tooling": "v22.22.1" },
      "plugins": ["magicLink"],
      "os": "macOS (Darwin 25.5.0)"
    }
    

    Which area(s) are affected? (Select all that apply)

    Backend

    Auth config (if applicable)

    import { betterAuth } from "better-auth"
    export const auth = betterAuth({
      emailAndPassword: {  
        enabled: true
      },
    });
    

    Additional context

    • Bisected away unrelated suspects: the hang reproduces on older framework code and is independent of DI scope changes -- it's purely 1.6.10 -> 1.6.11.
    • Reproduced deterministically in a real workerd integration test (request magic link -> capture token -> verify) that times out before the change.
    • Relevant code: @better-auth/core context/transaction.mjs runWithTransaction (opens adapter.transaction unconditionally) vs getCurrentAdapter (reuses the active trx); internalAdapter.consumeVerificationValue -> consumeOneWithHooks -> adapter consumeOne.
    • Worked around downstream by making the adapter's transaction reentrant (reuse the in-flight transaction via AsyncLocalStorage), which is also the correct semantics (nested transaction = one atomic unit) -- but it seems like runWithTransaction itself should reuse the active transaction.

    Interface

    No new interfaces are introduced.

    ADDITIONAL ADVICE

    Since you are given a git repository, you can use git commands to simplify your work. For example, if you made a mistake and want to revert your changes, you can use git checkout <file> to restore the file to its original state. You can also reset all changes in the repository using git reset --hard command. Additionally, you can use git stash and git stash pop to temporarily save your changes and restore them later. Finally, you can see the changes that you've made (which will be submitted when you call submit) using git status or git diff commands. However you don't need to use git add or git commit before submitting your changes. In fact, submit only submits currently unstaged changes because it uses git diff to get the changes that need to be submitted.

    CHECKLIST

    Before submitting your solution, please go over the following checklist and make sure you've done everything:

    • If an error was reported in the issue description, I have successfully reproduced it.
    • If an error was reported in the issue description, I have confirmed that my fix resolves the error.
    • I have thought about edge cases and made sure my fix handles them as well.
    • I have run existing tests in the repository that might have been affected by the change I made and confirmed that they pass. I want you to list every bullet from this checklist and write your reasoning for why you think you did it or didn't need to.

    Repository has been uploaded and your shell is currently at the repository root. Time to solve the issue!

    (Current directory: /better-auth, current file: none) bash-$

  2. Claude Assistant

    Let's use the edit special command directly on the file.

  3. Claude Assistant

    Now let's check the dist/build files if there are compiled .mjs versions that also need updating (since issue references context/transaction.mjs).

  4. Claude Assistant

    Now let's find getWithHooks and consumeOneWithHooks definition to see if it calls runWithTransaction again (the nested call causing deadlock).

  5. Claude Assistant

    Good, no circular import risk. Now let's implement the fix in factory.ts to make the consumeOne/incrementOne fallbacks reentrant-safe by using runWithTransaction + getCurrentAdapter.

  6. Claude Assistant

    The consumeOne fallback spans lines 1373-1410 (res = await withSpan through resultNeedsOutputTransform = false;). Let's verify exact boundaries.

  7. Claude Assistant

    Good, matches lines 1373-1409 (res=... through the closing );) plus 1410 resultNeedsOutputTransform. Let's replace lines 1373:1410.

  8. Claude Assistant

    Now let's do the same for incrementOne fallback.

  9. Claude Assistant

    Lines 1506-1553 (through resultNeedsOutputTransform = false; at 1553). Let's confirm exact line number for that.

  10. Claude Assistant

    Now replace lines 1506-1557 with the reentrant version.

  11. Claude Assistant

    Both fallback paths are now reentrant-safe. Now let's check for other similar adapter.transaction( direct calls elsewhere in the codebase that could suffer the same issue (e.g., in internal-adapter.ts line 124, sign-up.ts, organization plugin).

  12. Claude Assistant

    Let's build core package first and run these tests to check compile-correctness.

  13. Claude Assistant

    Good, pnpm works now. Let's check node_modules already installed.

  14. Claude Assistant

    Dependencies installed. Let's run type-check / tests for core package specifically.

  15. Claude Assistant

    Type checks pass. Let's view the final code around both edits to make sure formatting is fine.

  16. Claude Assistant

    Now let's run biome/format to fix indentation consistency and check the incrementOne block too.

  17. Claude Assistant

    Good. Now let's run biome format to normalize style.

  18. Claude Assistant

    Now let's run tests: core package's own tests plus adapter tests (memory-adapter, kysely-adapter, better-auth package tests related to magic-link/verification).

  19. Claude Assistant

    Confirmed - these failures are pre-existing environment issues unrelated to our change. Let's restore our changes.

  20. Claude Assistant

    All pass after building. Let's run the full core test suite again plus better-auth package tests relevant to magic-link and verification.

  21. Claude Assistant

    All 26 test files pass now. Let's also run the better-auth package tests related to magic-link, verification, and transaction to ensure end-to-end correctness.

  22. Claude Assistant

    Build succeeded. Now run relevant tests.

  23. Claude Assistant

    Need to build @better-auth/telemetry too.

  24. Claude Assistant

    All pass. Now let's write a quick reproduction script that simulates a connection-pool-limited custom adapter with a non-reentrant transaction, to confirm the deadlock is fixed.

  25. Claude Assistant

    Let's place the script within the repo so node resolves workspace packages.

  26. Claude Assistant

    With our fix and the freshly built core, the deadlock is resolved. Let's confirm it deadlocks WITHOUT our fix (stash again, rebuild, test).

  27. Claude Assistant

    Confirmed: without the fix, this deterministically deadlocks (mirroring the exact issue). With the fix, it resolves correctly. Let's restore our changes and rebuild.

  28. Claude Assistant

    Now let's remove the temporary repro script and clean up dist artifacts we don't want to commit (since dist wasn't part of repo originally, let's check git status).

  29. Claude Assistant

    All organization tests pass. Now let's run the sign-up tests and the full adapter test suites (kysely, memory, prisma, drizzle, mongo) plus scim.

  30. Claude Assistant

    All pass. Now let's run the full db/internal-adapter test suite and any transaction-related tests to be thorough.

  31. Claude Assistant

    Confirmed pre-existing failures unrelated to our changes (missing postgres connection & mongo-adapter build artifact from environment setup). Our fix is restored. Let's do a final full diff review.

  32. Claude Assistant

    Now let's find and update the incrementOne FIXME comment similarly.

  33. Claude Assistant

    Now let's re-run tsc and tests to confirm nothing broke with the comment edits.

  34. Claude Assistant

    All 415 tests pass across core, magic-link, email-verification, and internal-adapter. Let's do a final full diff review.

  35. Claude Assistant

    This is a clean, focused fix. Let's also check that the dist folders we generated during testing aren't tracked/staged, and clean up.

  36. Claude Assistant

    Clean — only the intended source changes are present. Let's run one more broad check with knip/lint to make sure nothing else complains, and a final biome lint check on modified files.

  37. Claude Assistant

    Let's fix import ordering.

  38. Claude Assistant

    Lint passes now. Let's rebuild core and rerun the full relevant test suite once more to make sure everything still works after the lint fix.

  39. Claude Assistant

    All 38 test files (644 tests) pass. Let's do a final review of the complete diff.

  40. Claude Assistant

    The changes are focused, minimal, and well-tested. Let me go through the checklist:

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.