Repository navigation
Description
TL;DR
Run the full-history benchmarks against the three Phase 2 synthetic datasets at Phase-2 cadence (1 s blocks), and publish per-profile latency numbers against the Phase-2 targets. All three datasets are complete 20,000-ledger sets (#867).
Phase 2 profile and targets
| SAC transfer | OZ token transfer | Soroswap swap | |
|---|---|---|---|
| Traffic | 5k TPS | 4k TPS | 1.5k TPS |
| Tx per ledger (TPL) | 5000 | 4000 | 1500 |
Block time 1 s (applied at replay — a dataset bakes in only tx model + TPL), retention 6 months.
Targets this run measures against:
- Ingest: Phase 2 defines no per-slice target for the meta-available-to-client path; the phase budget is 2.5 s end to end. The harness measures the ingest slice of that path; results are labeled as the ingest slice, never as the full path, and reported as a full distribution (composition with the read/stream slice is tracked separately).
- Reads: GetTransaction / GetEvents / GetLedgers / GetTransactions meet their SLAs at the Phase-2 TPL and block time.
Inputs
Set 2026-07-18-apply-load-20k (generated under #867):
| Dataset | TPL | Ledgers | Note |
|---|---|---|---|
sac-5000 |
5000 | 20,000 | |
custom_token-4000 |
4000 | 20,000 | shared with the Phase-1 campaign |
soroswap-1500 |
1500 | 20,000 | shared with the Phase-1 campaign |
Report
Per profile: p50 / p90 / p99 / max for the ingest slice at 1 s cadence, and read-endpoint latencies vs their SLAs. Use three-state verdicts: met / not met / not measured at this load. Where no per-slice target exists, report the distribution and mark the verdict not measured against a slice target.
Done when
All three Phase-2 datasets have published campaign results at 1 s cadence, with a summary table against the Phase-2 targets and explicit labeling of what was and was not measured.
Related
- Synthetic ledger datasets for Phases 1–3 ("golden packs") #867 — synthetic datasets (input)
- Full-history benchmark commands — ingest across all data types #771 / Full-history query benchmark commands — reads across all data types #856 — the harness commands this runs
- e2e benchmark campaigns workflow — dispatch, run on EC2, publish results #868 — campaign automation (this run may be manual; automation lands there)
- Read-endpoint SLAs under ingestion load #870 — read SLAs under ingestion load (deeper read-side scope)
- #TBD — Phase-1 campaign (template and shared packs; fill in once posted)
Metadata
Metadata
Assignees
Type
Projects
- StatusShow more project fieldsDone
Activity
marwen-abid commentedon Jul 24, 2026
Phase 2 (1 s blocks, p99 ≤ 400 ms):