Skip to content

docs(bench): retain scoped initial-validation measurements #126

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
rldyourmnd merged 1 commit into from
Oct 7, 2026
Merged

docs(bench): retain scoped initial-validation measurements #126

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
33 changes: 33 additions & 0 deletions docs/reports/rds-initial-path-validation-20261007.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -54,3 +54,36 @@ The full engine suite now runs in both supported CI test lanes. Linux/X11,
current-head CI, immutable release builds and installed qualification remain
required before rollout. This report does not close W3/W6/W10 or claim physical
interface, suspend, power-loss or causal input-to-pixel acceptance.

## Measurement harness increment

Current-source `rds-bench` measurements are retained in
[rds-initial-validation-bench-20261007.json](rds-initial-validation-bench-20261007.json).
Source: `378d636a93548e7a1b15c1b1c45bb1749619a9ea`, macOS arm64,
`transport-noq`, unoptimized stripped development profile, one run per case.
Synthetic same-host direct endpoints use temporary identities. No deployed host,
private configuration or foreign desktop is part of these measurements.

| Case | Samples or verified payload | Observed result |
|---|---|---|
| Iroh ping | 32 | p50 0.558ms; p95 0.765ms; p99/max 0.962ms |
| Noq ping | 32 | p50 0.428ms; p95 1.669ms; p99/max 1.859ms |
| Iroh receiver-verified transfer | 4MiB | 18.163MiB/s |
| Noq receiver-verified transfer | 4MiB | 25.547MiB/s |
| Noq controlled mid-transfer loss | 2MiB verified | 179 dropped datagrams; verified receipt; 1.770MiB/s |

Transfer includes bounded streaming generation/hashing and a receiver byte-count
plus BLAKE3 receipt. The recovery case imposes15% client-egress loss and50ms
delay for a1.503s window; it verifies2MiB rather than the4MiB ordinary transfer.
It proves completion under that controlled impairment, not standby validation,
physical-network recovery or user-visible desktop latency. One run and32 ping
samples cannot establish comparative superiority, reproducibility or broad
release acceptance. No milestone/checkpoint gate is closed by this increment.

The reproducible command family is
`rds-bench run --scenario CASE --backend BACKEND --iterations 32 --transfer-mib 4 --timeout-s 30`;
CASES are `ping`, `transfer` on both lanes and `recovery` on Noq. The versioned
[measurement method](../benchmark-transfer.md) remains authoritative. The full
corrected software CI passed both OS test lanes; the first Rust CodeQL attempt
ended during SARIF upload and its failed attempt remains retained. Final scanner
and installed acceptance are recorded independently in their deployment evidence.
Loading
Loading