fix: add serverless leak fix oneoff by MasterPtato · Pull Request #4318 · rivet-dev/rivet

MasterPtato · 2026-02-26T19:42:54Z

Description

Please include a summary of the changes and the related issue. Please also include relevant motivation and context.

Type of change

Bug fix (non-breaking change which fixes an issue)
New feature (non-breaking change which adds functionality)
Breaking change (fix or feature that would cause existing functionality to not work as expected)
This change requires a documentation update

How Has This Been Tested?

Please describe the tests that you ran to verify your changes.

Checklist:

My code follows the style guidelines of this project
I have performed a self-review of my code
I have commented my code, particularly in hard-to-understand areas
I have made corresponding changes to the documentation
My changes generate no new warnings
I have added tests that prove my fix is effective or that my feature works
New and existing unit tests pass locally with my changes

railway-app · 2026-02-26T19:43:09Z

🚅 Deployed to the rivet-pr-4318 environment in rivet-frontend

Service	Status	Web	Updated (UTC)
website	😴 Sleeping (View Logs)	Web	Feb 26, 2026 at 7:58 pm
frontend-cloud	😴 Sleeping (View Logs)	Web	Feb 26, 2026 at 7:53 pm
frontend-inspector	😴 Sleeping (View Logs)	Web	Feb 26, 2026 at 7:52 pm
ladle	❌ Build Failed (View Logs)	Web	Feb 26, 2026 at 7:46 pm
mcp-hub	✅ Success (View Logs)	Web	Feb 26, 2026 at 7:44 pm

MasterPtato · 2026-02-26T19:43:12Z

How to use the Graphite Merge Queue

Add the label merge-queue to this PR to add it to the merge queue.

You must have a Graphite account in order to use the merge queue. Sign up using this link.

_{An organization admin has enabled the Graphite Merge Queue in this repository.} _{Please do not merge from GitHub as this will restart CI on PRs being processed by the merge queue.}

This stack of pull requests is managed by Graphite. Learn more about stacking.

claude · 2026-02-26T19:47:45Z

PR Review: `fix: add serverless leak fix oneoff`

Status: Merged. Feedback below is for the record and future reference.

Overall this is a focused, well-scoped fix. The core logic (read orphaned slots from the unsubspaced location, atomically re-apply to the correct pegboard subspace in a single transaction) is correct. Below are issues found during review.

Medium Severity

1. dry_run + flip interaction is misleading (cli.rs)

When a user runs --dry-run -f, the logged slots value does not reflect the flip that would be applied, making dry-run validation inaccurate:

if dry_run {
    tracing::info!(?slots, "found lost slots");  // shows raw slots, ignores flip
} else {
    let slots = if flip { slots * -1 } else { slots };
    // ...
}

Suggestion: compute the (potentially-flipped) value before the if dry_run branch so both paths log the same effective value.

2. 5-second transaction timeout may be too tight

The operation uses StreamingMode::WantAll across the entire unsubspaced ServerlessDesiredSlotsKey space, applies atomic ops, and wraps everything in a 5-second tokio::time::timeout. FDB transactions also have a hard 5-second default limit. For a production environment with many namespaces/runners, this scan could time out silently before completing. Consider paginating the scan or at minimum adding a comment documenting this limitation.

Low Severity

3. slots * -1 can overflow for i64::MIN

In release builds, negating i64::MIN silently wraps back to i64::MIN. Prefer idiomatic Rust negation:

// Before
let slots = if flip { slots * -1 } else { slots };

// After
let slots = if flip { -slots } else { slots };

4. No success/completion message after transaction

The Ok(Ok(_)) => {} arm is silent. An operator has no way to distinguish "completed with 0 entries" from a quiet failure or timeout. Even a simple tracing::info!("reapply complete") would help.

5. No confirmation prompt before destructive write

Other destructive subcommands (e.g., Clear) require a --yes flag or interactive prompt. ReapplyLostServerless modifies production data directly when --dry-run is not set, relying entirely on operator discipline. Consider adding a --yes confirmation guard consistent with the rest of the CLI.

6. Missing #[command(name = "...")] on Oneoff variant

All other variants use explicit #[command(name = "...")] attributes (e.g., #[command(name = "cd")]). Oneoff uses clap's default snake-case derivation. Not a bug, but inconsistent with the established convention.

Non-issues / Good Choices

Correct cross-subspace transaction: Reading without subspace (where data is lost) and writing with the pegboard subspace in a single serializable transaction is the right FDB pattern.
rivet-types dependency promotion: Moving from dev-dependencies to dependencies is correct since it is now used in production code.
Unused protocol_version_entry removal (runner2.rs): The key was being read inside check_expired but never consumed after the try_join! destructuring — removing it is both correct cleanup and a small performance win (one fewer snapshot read per expiry check).
Unused import removals: All removals (Deserialize, MAX_BODY_SIZE) are accurate.
default_value_t = false cleanup: Correct — this is redundant for bool args in clap.
universaldb::prelude::* glob import: Acceptable for a prelude module; the restriction in CLAUDE.md is specific to anyhow.

pkg-pr-new · 2026-02-27T00:16:54Z

More templates

@rivetkit/virtual-websocket

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/virtual-websocket@4318

@rivetkit/cloudflare-workers

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/cloudflare-workers@4318

@rivetkit/framework-base

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/framework-base@4318

@rivetkit/next-js

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/next-js@4318

@rivetkit/react

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/react@4318

rivetkit

pnpm add https://pkg.pr.new/rivet-dev/rivet/rivetkit@4318

@rivetkit/sql-loader

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/sql-loader@4318

@rivetkit/sqlite-vfs

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/sqlite-vfs@4318

@rivetkit/traces

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/traces@4318

@rivetkit/workflow-engine

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/workflow-engine@4318

@rivetkit/engine-runner

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/engine-runner@4318

@rivetkit/engine-runner-protocol

pnpm add https://pkg.pr.new/rivet-dev/rivet/@rivetkit/engine-runner-protocol@4318

commit: 61a6e04

MasterPtato requested a review from NathanFlurry February 26, 2026 19:42

railway-app bot temporarily deployed to rivet-frontend / rivet-pr-4318 February 26, 2026 19:43 Destroyed

This was referenced Feb 26, 2026

fix(guard, runner): enforce http body size and ws msg size #4252

Merged

fix: various serverless bug fixes #4313

Merged

chore: fix tests #4314

Merged

MasterPtato mentioned this pull request Feb 26, 2026

fix(ups): upgrade nats, fix chunking, add chunking tests #4315

Merged

11 tasks

This was referenced Feb 26, 2026

fix: remove useless config params #4319

Merged

fix: always trigger serverless ended early error #4320

Merged

MasterPtato force-pushed the 02-25-chore_add_log_for_runners_with_no_alloc_idx branch from 5e58e8c to a9c0244 Compare February 27, 2026 00:12

MasterPtato force-pushed the 02-26-fix_add_serverless_leak_fix_oneoff branch from 8839aa5 to 61a6e04 Compare February 27, 2026 00:12

NathanFlurry approved these changes Feb 27, 2026

View reviewed changes

Base automatically changed from 02-25-chore_add_log_for_runners_with_no_alloc_idx to main February 27, 2026 00:31

fix: add serverless leak fix oneoff

5f5799b

NathanFlurry force-pushed the 02-26-fix_add_serverless_leak_fix_oneoff branch from 61a6e04 to 5f5799b Compare February 27, 2026 00:31

NathanFlurry merged commit f335eab into main Feb 27, 2026
6 of 10 checks passed

NathanFlurry deleted the 02-26-fix_add_serverless_leak_fix_oneoff branch February 27, 2026 00:31

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix: add serverless leak fix oneoff#4318

fix: add serverless leak fix oneoff#4318
NathanFlurry merged 1 commit intomainfrom
02-26-fix_add_serverless_leak_fix_oneoff

MasterPtato commented Feb 26, 2026

Uh oh!

railway-app bot commented Feb 26, 2026 •

edited

Loading

Uh oh!

MasterPtato commented Feb 26, 2026 •

edited

Loading

Uh oh!

claude bot commented Feb 26, 2026 •

edited

Loading

Uh oh!

pkg-pr-new bot commented Feb 27, 2026 •

edited

Loading

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

MasterPtato commented Feb 26, 2026

Description

Type of change

How Has This Been Tested?

Checklist:

Uh oh!

railway-app bot commented Feb 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

MasterPtato commented Feb 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

How to use the Graphite Merge Queue

Uh oh!

claude bot commented Feb 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Review: fix: add serverless leak fix oneoff

Medium Severity

Low Severity

Non-issues / Good Choices

Uh oh!

pkg-pr-new bot commented Feb 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

railway-app bot commented Feb 26, 2026 •

edited

Loading

MasterPtato commented Feb 26, 2026 •

edited

Loading

claude bot commented Feb 26, 2026 •

edited

Loading

PR Review: `fix: add serverless leak fix oneoff`

pkg-pr-new bot commented Feb 27, 2026 •

edited

Loading