node-accelerator v3.2 — na-diagnose --retrans (TCP retransmit root-cause) - #10
Merged
Conversation
`na-diagnose --retrans [--window N]` — докопаться до ПРИЧИНЫ retrans, а не только факта. За одно окно семплинга: rate, TX/RX-перекос, разбивка по типу (Timeouts/LostRetransmit/SackRecovery/SlowStart/BacklogDrop), хвост сокетов + распределение, реальный CC на проводе (bbr vs cubic), дропы TX-тракта (qdisc/softirq/NIC ring), accept-queue, TCP-фичи + вердикт (MSS-коллапс/qdisc/путь). Валидирован live: на реальной ноде корректно показал retrans 2.35% (TX-heavy, BBR на проводе, qdisc чист) и вердикт MSS-коллапс (min_snd_mss=48 + collapsed). ci: smoke прогоняет --retrans --window 5. docs: CHANGELOG v3.2 + README. Концепция вдохновлена наработками коллеги-оператора (по его просьбе — без имени). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Диагностика причин TCP-retransmits (read-only, поведение защиты/тюнинга не меняется).
🔬
na-diagnose --retrans [--window N]Докапывается до причины retrans, а не только до факта. За одно окно семплинга (по умолч. 20с):
out/in— download vs upload через ноду);TCPTimeouts/TCPLostRetransmit/TCPSackRecovery/TCPSlowStartRetrans/TCPBacklogDrop/ …;ss -tin: bbr vs cubic) — ловит «настройка bbr, а сокеты на cubic»;dropped/overlimits/requeues), softirqNET_RX/TX, NIC ring (ethtool -Sdrop/err);sack/dsack/frto/recovery/min_snd_mss/mtu_probing);Все инструменты опциональны (guard на
command -v ss/tc/ethtool), окно--window 5..300.✅ Проверка
Валидирован live на реальной ноде: корректно показал retrans 2.35% (TX-heavy, BBR на проводе, qdisc чист → причина = путь) и вердикт MSS-коллапс (
min_snd_mss=48+ collapsed sockets). CI smoke прогоняет--retrans --window 5.shellcheck --severity=errorчист.Концепция вдохновлена наработками коллеги-оператора (по его просьбе — без имени).
🤖 Generated with Claude Code