Design: TollGate capacity advertisement vs client measurement + reputation model

Đang mở
  • Có gì không ổn?
#313 2 bình luận 0 reaction 0 người được giao Xem trên GitHub

Maintainer thường phản hồi trong vòng 2 ngày

Chưa có ai nhận issue này.

Đánh giá

Độ khó
5/5
Thời gian dự kiến
Hơn một tuần
Mức phù hợp với người mới
25/100
Loại issue
Tính năng
Độ rõ ràng
Cần làm rõ
Mức độ hoạt động
Ít trao đổi
Công nghệ
go
Lĩnh vực
backend-api-design, networking

Hướng nghiên cứu

Start by reading the UpstreamManager, which the issue says already tracks bytes transferred during active sessions, and review #312 for the related discovery logger work. The issue leaves key decisions open about capacity measurement, reputation signals, and storage, so first clarify and agree on those choices. Done means an agreed design for how measurements and reputation will be collected and used.

Do mô hình lập chỉ mục viết ra từ nội dung của issue.

Mô tả

enhancement

Open Questions

1. Self-advertised capacity vs client-measured

Should a TollGate self-report its uplink speed, or should the client measure it?

Self-advertised (push model):

  • TollGate includes uplink_speed_mbps in its GET / advertisement response
  • Pro: client knows before committing to a connection
  • Con: nothing prevents lying — a TollGate could claim 100 Mbps on a 5 Mbps DSL line
  • Mitigation: client verifies after connecting and adjusts trust

Client-measured (pull model):

  • Client does a timed download from the TollGate backend after connecting
  • Pro: ground truth, can't be faked
  • Con: only available AFTER connection (too late for pre-connection scoring)
  • Con: small downloads are inaccurate for high-bandwidth links

Hybrid (likely best):

  • TollGate advertises expected capacity (from config or self-test)
  • Client measures actual throughput after connecting
  • Discrepancy logged → feeds reputation score
  • Future scoring weights measured > advertised
2. TollGate reputation

How do we track whether a TollGate is trustworthy and performs well?

Signals to collect:

  • Price accuracy: advertised price vs actual charge per step
  • Throughput delivery: advertised speed vs measured speed (ratio)
  • Uptime: what fraction of probes succeeded over time?
  • Payment reliability: did Cashu tokens work? Any meltdowns/errors?
  • Session stability: did sessions get cut prematurely?

Storage options:

  • Local JSONL per-TollGate (by npub) — simple, no infrastructure
  • Nostr reputation event (kind 1984 or custom) — shared across the network
  • Stake-based: TollGates lock sats as collateral, slashed for bad behavior

Trust bootstrapping:

  • New TollGate starts at neutral reputation (score 0)
  • Reputation accrues from client measurements over time
  • Social trust: follow npubs of known operators (profit_share identities)
  • Network effects: TollGates recommended by trusted TollGates get a boost
3. Measurement methodology

If we measure throughput from the client side:

  • HTTP GET to /: Already exists, returns advertisement (~500 bytes). Too small for throughput estimation.
  • Timed file download: Would need a dedicated endpoint serving a known-size payload (e.g., /speedtest?size=1000000). More accurate but adds backend complexity.
  • iperf3: Most accurate but requires installation on every TollGate.
  • Ongoing session throughput: The UpstreamManager already tracks bytes transferred during active sessions. This is "free" data — no extra probes needed. The session manager knows exactly how many bytes flowed through per second.

Proposal: Use ongoing session bytes/second as the primary throughput signal. The session manager already measures this for data-metered sessions. We just need to log it alongside the TollGate identity.

Related

  • #311 — WiFi AP selection algorithm roadmap
  • #312 — Discovery logger (Phase 1-2 implementation)
Ngôn ngữ chính
Go
Star
12
Fork
14
Merge trung bình
4 ngày 19 giờ
Pull request đã merge (30 ngày)
213

Chuẩn bị môi trường

Bắt đầu từ đâu

  1. Đọc hết issue, rồi đọc hướng dẫn đóng góp của dự án.
  2. Bình luận trên issue rằng bạn sẽ nhận — tránh hai người làm cùng một việc.
  3. Fork repository và làm thay đổi trên một nhánh.
  4. Mở pull request có tham chiếu số hiệu của issue.

Issue khác của OpenTollGate/tollgate-module-basic-go

Tất cả issue của OpenTollGate/tollgate-module-basic-go

Issue tương tự

Thêm issue về Go

Nhận issue mới trong hộp thư của bạn

Bản tóm tắt ngắn những issue GitHub phù hợp với người mới.