jedarden e4e9a16242 P1.6: Verify property + benchmark tests for router

Verify all acceptance criteria met:
- cargo bench -p miroir-core runs criterion benches
- cargo test runs proptest with 1024 cases (proptest.toml)
- CI includes cargo bench --no-run (miroir-ci.yaml:124)

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

2026-05-20 08:28:03 -04:00

2.3 KiB

Raw Blame History

P1.6 Property + Benchmark Tests Verification

Summary

Verified that all P1.6 acceptance criteria were already implemented in the codebase.

Acceptance Criteria Verification

1. `cargo bench -p miroir-core` runs all criterion benches and reports timing ✓

Files:

crates/miroir-core/benches/router_bench.rs - Rendezvous assignment benchmarks targeting §8 goals
crates/miroir-core/benches/merger_bench.rs - Merger benchmarks (1000 hits, 3 shards)

Benchmarks include:

shard_for_key single and batch (10K docs)
assign_shard_in_group single and all shards (64)
Full routing pipeline (10K docs through hash → shard → assign)
Varying shard counts (8, 16, 32, 64, 128, 256)
Varying node counts (2, 3, 4, 5, 8, 10)
Varying replication factors (1, 2, 3, 5)
Merger with varying hit counts (100, 500, 1000, 5000, 10000)
Merger with varying shard counts (1, 2, 3, 5, 10)
Pagination, facets, score preservation, degraded responses

2. `cargo test -p miroir-core` runs property tests with 1024 cases per property ✓

File: crates/miroir-core/tests/router_proptest.rs

Property tests (all with 1024 cases via proptest.toml):

Determinism (single and multiple runs)
Minimal reshuffling bounds on node add/remove
Uniformity (shards distribute evenly across nodes)
assign_shard_in_group returns exactly rf nodes
All returned nodes are from input set
No duplicate nodes in assignment
Score function uses different values for different inputs
shard_for_key determinism and valid range

Configuration: proptest.toml sets cases = 1024 for both default and CI profiles.

3. Phase 8 CI includes `cargo bench --no-run` to compile benches on every build ✓

File: k8s/argo-workflows/miroir-ci.yaml line 124

# Plan §8 Phase 1: compile benchmarks on every build (without running)
cargo bench --no-run

Test Results

All 12 proptest properties pass:

prop_assign_no_duplicates
prop_assign_nodes_from_input
prop_assign_returns_rf_nodes
prop_determinism
prop_determinism_multiple_runs
prop_reshuffle_bound_on_add
prop_reshuffle_bound_on_remove
prop_score_different_inputs
prop_shard_for_key_determinism
prop_shard_for_key_valid_range
prop_uniformity
regression_tests::test_shard_for_key_known_values

All benchmarks compile successfully.

2.3 KiB Raw Blame History