perf(trace-serializer): pre-allocate serialization buffer#1949
Merged
Conversation
# Motivation When serializing spans, we start from a zero sized buffer. Pre-allocate the buffer to prevent small re-allocs. # Changes * Save the length of the payload we serialised * When creating a new buffer allocate capacity for the length of the previous payload * Also add a 1KB minimum buffer size.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 5a2f4c701e
ℹ️ About Codex in GitHub
Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".
Comment on lines
+105
to
+108
| let capacity = self | ||
| .previous_serialised_len | ||
| .load(Ordering::Relaxed) | ||
| .max(MIN_BUFFER_CAPACITY); |
There was a problem hiding this comment.
In the production agent path I checked, send_trace_chunks_inner constructs a fresh TraceSerializer immediately before the single prepare_traces_payload call (libdd-data-pipeline/src/trace_exporter/mod.rs:619-623), so this load always sees the constructor's MIN_BUFFER_CAPACITY value rather than the previous payload size. That means batches larger than 1KB still grow from a 1KB allocation on every send, and the stored length is discarded when the local serializer is dropped; to get the intended pre-allocation, the serializer state needs to live across send calls.
Useful? React with 👍 / 👎.
Contributor
There was a problem hiding this comment.
If this is true, maybe previous_serialized_len should be a static ? Or would the scope be too large?
Contributor
Author
There was a problem hiding this comment.
It is true, I just refactored the TraceSerializer and stored it in the TraceExporter, this way it persists across sends
Contributor
📚 Documentation Check Results📦
|
Contributor
Clippy Allow Annotation ReportComparing clippy allow annotations between branches:
Summary by Rule
Annotation Counts by File
Annotation Stats by Crate
About This ReportThis report tracks Clippy allow annotations for specific rules, showing how they've changed in this PR. Decreasing the number of these annotations generally improves code quality. |
Contributor
🔒 Cargo Deny Results📦
|
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #1949 +/- ##
==========================================
- Coverage 71.85% 71.81% -0.05%
==========================================
Files 434 434
Lines 70454 70621 +167
==========================================
+ Hits 50624 50713 +89
- Misses 19830 19908 +78
🚀 New features to boost your workflow:
|
Contributor
🎉 All green!❄️ No new flaky tests detected 🎯 Code Coverage (details) 🔗 Commit SHA: da0086a | Docs | Datadog PR Page | Give us feedback! |
Contributor
Artifact Size Benchmark Reportaarch64-alpine-linux-musl
aarch64-unknown-linux-gnu
libdatadog-x64-windows
libdatadog-x86-windows
x86_64-alpine-linux-musl
x86_64-unknown-linux-gnu
|
yannham
approved these changes
May 7, 2026
Comment on lines
+116
to
+119
| # rmp ={ path = "../../../../../projects/msgpack-rust/rmp" } | ||
| # rmp-serde ={ path = "../../../../../projects/msgpack-rust/rmp-serde" } | ||
| # rmp ={ path = "/projects/msgpack-rust/rmp" } | ||
| # rmp-serde ={ path = "/projects/msgpack-rust/rmp-serde" } |
Contributor
Author
There was a problem hiding this comment.
yes, from a previous experiment
iunanua
added a commit
that referenced
this pull request
May 18, 2026
# Release proposal for libdd-data-pipeline and its dependencies This PR contains version bumps based on public API changes and commits since last release. ## libdd-dogstatsd-client **Next version:** `3.0.0` **Semver bump:** `major` **Tag:** `libdd-dogstatsd-client-v3.0.0` ###⚠️ major bump forced due to: - `libdd-common`: ^3.0.2 → ^4.1.0 ### Commits - fix(crypto): gate libdd-common TLS features in remaining internal crates + add CI guard (#1943) ## libdd-trace-obfuscation **Next version:** `3.0.0` **Semver bump:** `major` **Tag:** `libdd-trace-obfuscation-v3.0.0` ###⚠️ major bump forced due to: - `libdd-common`: ^3.0.2 → ^4.1.0 - `libdd-trace-utils`: ^3.0.1 → ^4.0.0 ### Commits - feat!: integrate obfuscation to the stats exporter [APMSP-2764] (#1819) - feat!: added regex-lite feature (#1939) - chore: clippy (#1889) - fix(crypto): gate libdd-common TLS features in obfuscation and capabilities-impl (#1872) - feat(obfuscation)!: feature parity on span obfuscation [APMSP-2671] (#1788) - feat(obfuscation/sql): feature parity on sql obfuscation [APMSP-2667] (#1708) ## libdd-trace-stats **Next version:** `3.0.0` **Semver bump:** `major` **Tag:** `libdd-trace-stats-v3.0.0` ###⚠️ major bump forced due to: - `libdd-trace-utils`: ^3.0.1 → ^4.0.0 ### Commits - perf: pre-compute string messagepack encoding (#1948) - feat!: integrate obfuscation to the stats exporter [APMSP-2764] (#1819) - feat(capablities)!: sleep & spawn capabilities (#1873) - feat: use ip quantization when aggregating peer tags for trace stats (#1944) - fix(crypto): gate libdd-common TLS features in remaining internal crates + add CI guard (#1943) - feat(shared_runtime)!: allow worker to be stopped after fork (#1893) - feat(sidecar)!: Add stats computation via SHM (#1821) - feat(stats): propagate service source from span meta to client stats payload (#1803) - fix(stats): align with css spec (#1790) ## libdd-data-pipeline **Next version:** `4.0.0` **Semver bump:** `major` **Tag:** `libdd-data-pipeline-v4.0.0` ###⚠️ major bump forced due to: - `libdd-common`: ^3.0.2 → ^4.1.0 - `libdd-telemetry`: ^4.0.0 → ^5.0.0 - `libdd-trace-utils`: ^3.0.1 → ^4.0.0 ### Commits - fix(libdd-telemetry): restore previous Cargo.toml version (#1993) - fix(data-pipeline): remove default-features from of trace-obfuscation (#1981) - fix(trace_exporter: shared_runtime): unwrap_or being eager is not good (#1983) - perf: pre-compute string messagepack encoding (#1948) - feat!: integrate obfuscation to the stats exporter [APMSP-2764] (#1819) - feat(capablities)!: sleep & spawn capabilities (#1873) - fix(telemetry): avoid trigger loop in telemetry worker (#1950) - feat(telemetry)!: include dependencies and integrations in app-extended-heartbeat (#1962) - perf(trace-serializer): pre-allocate serialization buffer (#1949) - feat!: added regex-lite feature (#1939) - fix(crypto): gate libdd-common TLS features in remaining internal crates + add CI guard (#1943) - feat(telemetry): add session id support to trace export (#1822) - fix(path): missing bench path in data-pipeline (#1907) - feat(data-pipeline): port dd-trace-rs trace buffer implementation (#1826) - feat(info_fetcher): add timeout to info fetcher (#1890) - feat(shared_runtime)!: allow worker to be stopped after fork (#1893) - feat(sidecar)!: Add stats computation via SHM (#1821) - ci(libdd-shared-runtime): downgrade version so publish workflow succeeds (#1870) - feat(runtime)!: add shared runtime (#1602) - ci: compilation of libdd-data-pipeline to wasm32 (#1830) - feat(capabilities)!: trait architecture http (#1555) - feat(otel): add support for OTLP trace export (#1641) - fix(stats): align with css spec (#1790) --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: iunanua <18325288+iunanua@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Motivation
When serializing spans, we start from a zero sized buffer. Pre-allocate the buffer to prevent small re-allocs.
Changes