Skip to content

deps: bump vgi-rpc from 0.16.0 to 0.23.3 - #19

Open
dependabot[bot] wants to merge 1 commit into
mainfrom
dependabot/cargo/vgi-rpc-0.23.3
Open

deps: bump vgi-rpc from 0.16.0 to 0.23.3#19
dependabot[bot] wants to merge 1 commit into
mainfrom
dependabot/cargo/vgi-rpc-0.23.3

Conversation

@dependabot

@dependabot dependabot Bot commented on behalf of github Sep 1, 2026

Copy link
Copy Markdown

Bumps vgi-rpc from 0.16.0 to 0.23.3.

Release notes

Sourced from vgi-rpc's releases.

v0.23.3

Fixed

  • Subprocess clients can enforce a monotonic response deadline per RPC. An expired call poisons the connection so late pipe bytes cannot desynchronize a later call or return the client to a pool. Unix kills and reaps the worker's private process group before joining its reader. Other platforms kill and reap the direct child, then bound the reader join and detach it if a descendant retained stdout.
  • Synchronous unary, stream-init, producer, exchange, cancellation, state decode, and state-encode callbacks no longer occupy Tokio HTTP workers on a multi-thread runtime. Slow worker code now yields async request capacity via block_in_place, so unrelated requests and post-timeout recovery continue immediately even on a one-worker runtime.

v0.23.2

Performance

  • Non-dictionary shared-memory batches now decode directly from mmap-backed Arrow buffers while retaining stock crates.io arrow-rs. Allocator slots remain owned until the final array view is dropped, avoiding both the inbound payload copy and premature slot reuse.
  • Shared-memory serialization now uses the direct single-row LargeBinary writer introduced in 0.23.1, and allocator entries preserve Arrow IPC alignment between consecutive batches. On the canonical Linux benchmark, 16 MiB echo latency fell from 11.91 ms to 9.56 ms (3.27 GiB/s), making shared memory 14% faster than the 11.12 ms Unix-socket result.

v0.23.1

Added

  • LargeBytesBuffer provides an immutable, reference-counted LargeBinary handler value. Reading retains a slice of the inbound Arrow allocation and returning it transfers that buffer into the response array without copying.

Performance

  • LargeBytes(Vec<u8>) response construction now transfers the vector's allocation directly into Arrow instead of copying it through a builder.
  • Single-row LargeBinary batches write their value buffer directly to the destination pipe, socket, HTTP body, or shared-memory serializer. This avoids the additional payload-sized staging vector created by arrow-rs's general IPC encoder while retaining canonical Arrow IPC framing and custom metadata.
  • The conformance worker's echo_large_binary method uses the zero-copy type. Controlled local 16 MiB echo latency improved from 11.36 ms to 7.52 ms over Unix sockets and from 14.13 ms to 10.07 ms over loopback TCP, with small-call latency unchanged.

... (truncated)

Changelog

Sourced from vgi-rpc's changelog.

[0.23.3] — 2026-08-27

Fixed

  • Subprocess clients can enforce a monotonic response deadline per RPC. An expired call poisons the connection so late pipe bytes cannot desynchronize a later call or return the client to a pool. Unix kills and reaps the worker's private process group before joining its reader. Other platforms kill and reap the direct child, then bound the reader join and detach it if a descendant retained stdout.
  • Synchronous unary, stream-init, producer, exchange, cancellation, state decode, and state-encode callbacks no longer occupy Tokio HTTP workers on a multi-thread runtime. Slow worker code now yields async request capacity via block_in_place, so unrelated requests and post-timeout recovery continue immediately even on a one-worker runtime.

[0.23.2] — 2026-08-24

Performance

  • Non-dictionary shared-memory batches now decode directly from mmap-backed Arrow buffers while retaining stock crates.io arrow-rs. Allocator slots remain owned until the final array view is dropped, avoiding both the inbound payload copy and premature slot reuse.
  • Shared-memory serialization now uses the direct single-row LargeBinary writer introduced in 0.23.1, and allocator entries preserve Arrow IPC alignment between consecutive batches. On the canonical Linux benchmark, 16 MiB echo latency fell from 11.91 ms to 9.56 ms (3.27 GiB/s), making shared memory 14% faster than the 11.12 ms Unix-socket result.

[0.23.1] — 2026-08-24

Added

  • LargeBytesBuffer provides an immutable, reference-counted LargeBinary handler value. Reading retains a slice of the inbound Arrow allocation and returning it transfers that buffer into the response array without copying.

Performance

  • LargeBytes(Vec<u8>) response construction now transfers the vector's allocation directly into Arrow instead of copying it through a builder.
  • Single-row LargeBinary batches write their value buffer directly to the destination pipe, socket, HTTP body, or shared-memory serializer. This avoids the additional payload-sized staging vector created by arrow-rs's general IPC encoder while retaining canonical Arrow IPC framing and custom metadata.
  • The conformance worker's echo_large_binary method uses the zero-copy type. Controlled local 16 MiB echo latency improved from 11.36 ms to 7.52 ms over Unix sockets and from 14.13 ms to 10.07 ms over loopback TCP, with small-call latency unchanged.

... (truncated)

Commits
  • 4fead3a release: prepare v0.23.3
  • b4e528f Add tagline and badges to README header
  • ae93a4e fix(client): bound non-Unix reader teardown
  • b340e28 fix(client): bound subprocess response waits
  • c093504 fix(http): keep async workers responsive
  • eaba89b release: prepare v0.23.2
  • 8e246a2 shm: restore zero-copy Rust payload decoding
  • c0a61c5 perf: remove large binary response copies
  • ad9cef2 ci: test conformance against Python main
  • c8a2b9f feat(protocol): complete lock-step and gzip conformance
  • Additional commits viewable in compare view

Dependabot compatibility score

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.


Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

  • @dependabot rebase will rebase this PR
  • @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
  • @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
  • @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
  • @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
  • @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)

Bumps [vgi-rpc](https://github.com/Query-farm/vgi-rpc-rust) from 0.16.0 to 0.23.3.
- [Release notes](https://github.com/Query-farm/vgi-rpc-rust/releases)
- [Changelog](https://github.com/Query-farm/vgi-rpc-rust/blob/main/CHANGELOG.md)
- [Commits](Query-farm/vgi-rpc-rust@v0.16.0...v0.23.3)

---
updated-dependencies:
- dependency-name: vgi-rpc
  dependency-version: 0.23.3
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
@dependabot dependabot Bot added dependencies Pull requests that update a dependency file rust Pull requests that update rust code labels Sep 1, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

dependencies Pull requests that update a dependency file rust Pull requests that update rust code

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants