v0.15.0 milestone snapshot (condensed) · flight v0.15.0@sha256:30c2b10c (multi-arch INDEX) · connector 0.15.0 · Trino 481 · Cassandra 5.0 RF=3 · 3× i4i.xlarge · ~1.93M partitions/node, 2 SSTable gens · 2026-07-17
| Metric | R11b | R12 | 0.15 (snap) | |
|---|---|---|---|---|
| warm throughput | ~34 qps @8-thr | ~33 qps @8-thr | ~39 qps @32-thr | parity+ |
| warm p50 / p99 | 227 / 366ms | 242 / 363ms | 798 / 1366ms @32-thr | higher concurrency |
| count(*) rows | 1,939,286 | 1,947,775 | 1,927,467 | dataset-dependent |
| count(*) wall | 66.2s | 61.1s | ~60s | parity |
| idle RSS | 3–4 Mi | 3 Mi | 4–5 Mi | parity |
| peak RSS | 270–391Mi | ~603MB @80-thr | ~310Mi @80-thr | lower |
| do_get error rate | 2.3% | 1.2% | 0.89% | improving |
| snapshot grace-sweep | — | query-triggered (738 held) | background (660→6, no query) | FIXED |
| OOMKills / restarts | 0 / 0 | 0 / 0 | 0 / 0 | clean |
Note: p50/p99 are not directly comparable — R11b/R12 measured 8-thread load, this run 32-thread. At equal concurrency the numbers track; the higher latency here is queuing at 4× the threads against the same throughput ceiling.
| Goal | Target | 0.15 | |
|---|---|---|---|
| B1 warm point-read floor | ≤300ms | ~800ms p50 @32-thr (incl. ~2.2s JDBC floor amortized) | below goal at 32-thr |
| B1 warm scan-lite floor | ≤500ms | LIMIT-class within steady p50 | report |
| B2 throughput | ≥100 qps @32-thr | ~39 qps | NOT met — fan-out skew is the bottleneck |
| B3 full-scan trajectory | 66s→10s | ~60s | on old trajectory (this run didn't target scan tuning) |
| B4 cold start | ≤3s / idle ≤16Mi | 2.2s / 4–5Mi | MET |
The Lane-B ladder is progress-tracking, not pass/fail. The one clear gap — B2 throughput — traces directly to the fan-out skew finding: a single pod is the ceiling. Cold-start (B4) is comfortably met.