Weekly Performance Report - 2026-04-06 (v2.6.37)

[github-actions]

⚡ Performance Report — Week of 2026-04-06 (v2.6.37)

Automated analysis by the CLI Performance Monitor. Rust binary not present in CI, so measurements reflect the Python bootstrap layer (rust_bridge.py) which is the overhead before the Rust binary execs.

---

📦 Package & Bootstrap Startup

Metric	Time	Budget	Status
import azlin (package __init__)	24.6ms avg	<50ms	✅ Excellent
import azlin.rust_bridge (full module)	121.5ms avg	<150ms	✅ Good
P95 bootstrap time	122.6ms	<150ms	✅ Within budget
Min bootstrap time	120.2ms	—	—
Max bootstrap time	122.6ms	—	—

Dominant cost: urllib.request import (~111ms)

The rust_bridge.py module eagerly imports urllib.request at module load time (line 8), which accounts for ~92% of the 121ms bootstrap cost. This is pure stdlib overhead — no IO is performed, but the import itself is expensive.

---

🏗️ Architecture Note

Azlin v2.3.0+ is a Rust-first CLI with a thin Python bootstrap:

python3 -c "from azlin.rust_bridge import entry; entry()"
    └─ rust_bridge.py (121ms)  ← measured in this report
        └─ exec() Rust binary  ← not present in CI environment
            └─ Clap parse (typ. 5-10ms)
            └─ Command execution

The Rust binary itself (once installed) provides 75–85x faster startup than the original Python CLI (per CHANGELOG v2.3.0-rust). Live command benchmarks (azlin list, azlin connect, etc.) require the Rust binary and Azure access, which are unavailable in CI.

---

📊 Bootstrap Cost Breakdown (stdlib imports)

Import	Cost	Notes
urllib.request	~111ms	Largest single cost
subprocess	~23ms	Already included in combined
platform	~27ms	Already included in combined
shutil	~22ms	Already included in combined
pathlib	~17ms	Already included in combined
All combined	~117ms	Python deduplicates shared stdlib

---

🔍 Optimization Opportunity

Lazy urllib.request import — This is the highest-impact optimization available in the Python bootstrap layer:

# Current (eager, at module level):
import urllib.request

# Recommended (lazy, inside download function only):
def _download_binary(url: str, dest: Path) -> None:
    import urllib.request  # Only imported when actually needed
    ...

Estimated impact: ~90ms reduction in bootstrap time (from ~121ms → ~30ms) for users who already have the Rust binary installed (the common case). The download path is only taken on first install or self-update, not on every CLI invocation.

Affected file: src/azlin/rust_bridge.py line 8

---

📈 Performance Baselines (Current)

Metric	Current	Target	Status
Python bootstrap (P95)	122.6ms	<150ms	✅
Package __init__ import	24.6ms	<50ms	✅
Rust binary startup (expected)	~5-10ms	<100ms	✅ (per v2.3.0 notes)
azlin list (Rust, with cache)	N/A in CI	<200ms	—
azlin list (Rust, cold)	N/A in CI	<500ms	—

---

✅ Summary

• No regressions detected vs v2.6.36 (only change was a version bump)
• Bootstrap layer is within all performance budgets
• One optimization opportunity: lazy urllib.request import could save ~90ms on every invocation for existing installs
• Rust binary benchmarks (list, connect, create, delete) require an Azure-connected environment with the binary installed — these cannot be measured in CI

---

📋 Recommendations

1. Low-effort win: Lazy-import urllib.request inside _download_binary() and _build_from_source() only — saves ~90ms per invocation for users with the Rust binary already installed
2. Track Rust binary perf separately: Add a CI step that downloads the published binary and benchmarks --help, list --help, etc. (no Azure needed for help/version commands)
3. Add azlin version timing: The version subcommand requires no Azure and gives a clean Rust startup benchmark

Report generated by CLI Performance Monitor workflow on 2026-04-06 for azlin v2.6.37

Generated by CLI Performance Monitor