fix(metrics): prevent unbounded histogram memory growth

[moonming]

Summary

• periodically drain pending Prometheus histogram samples when AISIX owns a low-level recorder
• stop the upkeep task through the existing shared shutdown channel
• verify periodic upkeep scheduling and shutdown cancellation with paused Tokio time

Root cause

AISIX builds metrics-exporter-prometheus with PrometheusBuilder::build_recorder(). That low-level API does not spawn the exporter's upkeep task. Without Prometheus scrapes, each request leaves pending histogram samples in the recorder, so RSS grows approximately linearly with request count.

This is the memory issue tracked from api7/AISIX-Cloud#1199. It is independent of the allocator: local glibc and jemalloc runs both showed roughly constant retained bytes per request, and an unchanged binary stopped growing when /metrics was scraped periodically.

Benchmark

GetBusbar/benchmarking, OpenAI-to-OpenAI cell, ARM64 Colima VM, 8 vCPU / 6.8 GiB, gateway pinned to 4 cores, 300-second memory load:

Metric	Before	After
Idle RSS	67.20 MiB	67.26 MiB
Load-window RSS	553.04 MiB	90.81 MiB
Peak RSS	701.76 MiB	91.64 MiB
Recovered RSS	701.76 MiB	89.77 MiB
Growth	+118.42 MiB/min	-1.45 MiB/min
Plateau	false	true
Added latency p99	225 us	197 us
RPS under 10 ms p99	33,527	35,922

A second post-fix run measured 89.67 MiB peak, 87.34 MiB recovered, and -1.66 MiB/min growth.

Independent audit

• MEDIUM: the original histogram test could pass because render() also drains pending samples, so it did not prove the scheduler lifecycle. Resolved by factoring the upkeep loop and testing periodic execution plus cancellation with paused Tokio time.

Validation

• cargo test -p aisix-obs — 147 passed, 5 ignored real-cloud smoke tests
• cargo test -p aisix-server — 87 passed
• cargo clippy -p aisix-obs -p aisix-server --all-targets -- -D warnings
• cargo fmt --all -- --check
• GetBusbar OpenAI-to-OpenAI full memory protocol, two post-fix runs

[coderabbitai]

Warning

Review limit reached

@moonming, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 38 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 570cd7ce-22b7-4de0-9b0d-9169096e30d6

📥 Commits

Reviewing files that changed from the base of the PR and between 22b8353 and 1f3cfe6.

📒 Files selected for processing (3)

• crates/aisix-obs/src/metrics.rs
• crates/aisix-server/Cargo.toml
• crates/aisix-server/src/main.rs

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}