fix(sqlite-native): replace stderr sql trace logging

[NathanFlurry]

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

[NathanFlurry]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• fix(sqlite-native): replace stderr sql trace logging #4629 👈 (View in Graphite)
• fix(pegboard): repair follow-up build errors #4628
• docs(pegboard-runner): explain alloc idx eviction invariant #4626
• fix(pegboard): isolate runner config dc lookup failures #4625 : 1 other dependent PR (#4630 )
• docs(pegboard): note runner config upsert split write risk #4624
• fix(pegboard): unpack global runner config keys correctly #4623
• fix(pegboard-gateway): enforce tunnel message state #4622
• fix(pegboard-runner): clear terminal tunnel routes #4621
• fix(pegboard): skip protocol version keys in runner pool backfill #4620
• fix(pegboard): create missing normal runner configs #4619
• fix(epoxy): reuse cached optimistic value #4618
• main

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

[claude]

PR Review: fix(sqlite-native): replace stderr sql trace logging

This is a clean, targeted fix that correctly addresses a CLAUDE.md style violation (eprintln! → tracing::trace!). The changes are straightforward and all four call sites are handled consistently.

Positives

• Correctly replaces eprintln! with tracing::trace! per the CLAUDE.md rule: "Never use eprintln! or println! for logging in Rust code. Always use tracing macros."
• Good use of structured logging fields (op = %op_name, duration_us) instead of formatting into the message string.
• Log message follows the lowercase convention from CLAUDE.md.
• Appropriate log level: trace is correct for high-frequency, verbose diagnostic output.

Observations

Performance: std::env::var on every KV operation

The RIVET_TRACE_SQL guard calls std::env::var() on every single KV operation, which acquires a mutex on the process environment each time. Given that KV operations are on the hot path for SQLite (called for every read/write/delete), this could add non-trivial overhead in workloads with many small SQLite operations. Consider caching it once at startup:

use std::sync::LazyLock;
static TRACE_SQL: LazyLock<bool> = LazyLock::new(|| std::env::var("RIVET_TRACE_SQL").is_ok());

// Then in each method:
if *TRACE_SQL {
    tracing::trace!(...);
}

This is pre-existing behavior that the PR inherited, so it is not a blocker — but worth addressing in a follow-up if not here.

Duplicate log lines when RIVET_TRACE_SQL is set

When the env var is set, each KV op emits two log lines: the new tracing::trace! ("sql trace kv round-trip") and the existing tracing::debug! ("kv round-trip") just below it, with identical fields. This is also pre-existing, but now that the trace log is a proper tracing call, it may be worth consolidating — if debug always fires, does the conditional trace duplicate add enough signal?

Minor: the as u64 cast

elapsed.as_micros() returns u128. The cast to u64 matches what the existing tracing::debug! already does, so this is consistent — just noting it is technically lossy (though only for durations exceeding ~584,000 years).

Summary

The core fix is correct and aligns with CLAUDE.md. The performance concern with repeated std::env::var calls on the hot path is the one thing worth addressing — either here or in a follow-up. Otherwise, looks good.