fix(query): union instances_matching into ASAP-tier sid resolution (schema-retirement #5 step 1)

[zzylol]

Summary

First step of #272 (schema-retirement #5) — the query-path-side sid resolution fix that unblocks MVP demo Axis C (#271).

execute_range_promql_modern and the QueryEngine::execute() trait impl resolved ASAP-tier candidates → sids via sids_for_policy(fp) only, on the assumption that every production sid registration would populate policy_fp (per the comment a prior PR added when removing the legacy instances_matching fallback). That assumption breaks for sketches arriving via OTLP: derive_sketch_policy_fp only returns Some(fp) when a streaming-config policy's grouping_labels exactly matches the wire DP's group_by_keys. When the agent emits sketches with the full wire-attr set (no upstream attributes/keep — the common case when the controller's OpAMP-pushed runtime config doesn't reach the agent, see ASAPCollector#381), all sids land in the catalog with policy_fp = UNSET, and sids_for_policy(streaming_config_fp) returns empty even though the data is sitting in the SketchStore.

Fix: union the instances_matching(metric, group_by_keys) catalog walk into the sid set. instances_matching is the more general primitive — returns sids whose group_by_keys is a SUPERSET of the candidate's asked grouping. That subsumes the policy-fp reverse-index hit (ExactAgg sids minted via ingest_precompute_for_agg_config have group_by_keys == streaming_config.grouping_labels so they satisfy the subset check) AND covers the full-attr sketch case the prior path missed.

End-to-end verification

Single-node MVP smoke test (/mydata/mvp-smoke-test/):

Pre-fix:

$ curl --data-urlencode 'query=quantile_over_time(0.99, http_requests_total_latency_ms[5m])' \
       http://localhost:19091/api/v1/query
{"data":null,"error":"No result for query","errorType":"bad_data","infos":["data_source: asap_query"],"status":"error"}

Post-fix:

$ curl --data-urlencode 'query=quantile_over_time(0.99, http_requests_total_latency_ms[5m])' \
       http://localhost:19091/api/v1/query
{"accuracy":{"delta":0.0,"epsilon":0.01,"kind":"relative_quantile"},
 "data":{"result":[{"metric":{"zone":""},
                    "value":[1779043268.182,"90.93548893834691"]}],
         "resultType":"vector"},
 "infos":["accuracy: ε=0.01, δ=0, kind=relative_quantile",
          "precompute_window: [1779042780000, 1779042810000) ms (width 30000 ms)",
          "data_source: asap_query"]}

DDSketch p99 = 90.9ms with the right accuracy envelope. Axis D (51 sids registered) had been green for two weeks; this PR closes the loop on Axis C.

Test plan

• cargo test -p data_plane --lib → 756 pass / 0 fail / 5 ignored
• cargo test -p control_plane --lib → 691 pass / 0 fail
• New regression test full_attr_sketch_sid_findable_via_subset_grouping in asap_tier_classify_tests pins the smoke-test scenario at unit level: a sketch sid with policy_fp = UNSET and full wire-attr group_by_keys must be findable by a query whose group_by_keys is a subset
• End-to-end smoke test via bash /mydata/mvp-smoke-test/run_smoke.sh — Axis C quantile_over_time returns a real DDSketch quantile

Scope / what's still left

This is step 1 of #272. Still TODO:

• data_plane/src/drivers/ingest/otel.rs:561,640 — ingest bucketing rekey from (aggregation_id, group_key) to sid
• data_plane/src/precompute_engine/{worker,output_sink}.rs — WorkerMessage::AccumulatorInput shape change
• data_plane/src/storage_engines/sketch_db/backfill/processor.rs — 6 sites
• The legacy handle_query / execute_context path (engine.rs:945-1127) still routes via agg_id and emits the "No precomputed outputs found for metric: X, aggregation_id: Y" warning — its retirement is the natural next step
• Re-enable the two ignored cross-reconfigure dispatch tests in tests/schema_timeline_dispatch_tests.rs
• Drop pub fn aggregation_id() on AggregationConfig once all consumers are gone

Each is a separate PR per the staging plan in #272.

Related: #271 (MVP demo axis C) — this PR is sister fix path 2 from the issue body.

🤖 Generated with Claude Code