Repository navigation
feat: add quic-tcp-ec2-integration spec and ROADMAP items 37-38 - #122
Merged
Merged
Conversation
Synthetic Performance Results — Graviton (run)Commit: ✅ synthetic UDP socket bound to 10.0.0.1:9000 (MAC: 02:00:00:00:00:01) Synthetic UDP Performance ResultsMeasures framework overhead: sync IPv4 Baseline
IPv6
IPv6 vs IPv4 Comparison (sync path)
IPv4 avg sync/async ratio: 1.0x, worst: 1.1x | IPv6 vs IPv4 worst ratio: 1.23x (OK)
|
Integration Test Failure — Graviton (Run 33505065517)Branch: failure-summary.json{
"failed_step": "deploy_infrastructure",
"error": "Infrastructure deployment failed",
"exit_code": 2,
"timestamp": "2026-09-01T12:09:19.887053Z",
"sender_instance_id": "i-0b8664dad5014cf4d",
"receiver_instance_id": "i-09d42a02c9cf7bd55",
"commit": "1d308388331a74d510f3271cc58febf6c492337a",
"run_url": "https://github.com/gspivey/dpdk-stdlib-rust/actions/runs/33505065517"
}```
### receiver-console-output.log (65117 bytes, last 80 lines)[ 2332.997695] __mutex_lock.constprop.0+0x2cc/0x91c (console output unavailable) |
❌ Integration Tests Failed — Graviton (run)Branch: Test ResultsApplication Logs (last 20 lines) |
TCP Synthetic Performance Results (run)Commit: TCP Synthetic Performance ResultsMeasures TCP engine framework overhead using a mock PacketBackend.
conn_est=1881ns, throughput=4412.1Mbps, tick=709ns |
Synthetic Performance Results (run)Commit: ✅ synthetic UDP socket bound to 10.0.0.1:9000 (MAC: 02:00:00:00:00:01) Synthetic UDP Performance ResultsMeasures framework overhead: sync IPv4 Baseline
IPv6
IPv6 vs IPv4 Comparison (sync path)
IPv4 avg sync/async ratio: 0.9x, worst: 1.0x | IPv6 vs IPv4 worst ratio: 1.32x (OK)
|
[CI] Stage: DeployInfrastructure ready.
|
Integration Tests —
|
| Tier | Test | Result | Time | Reason |
|---|---|---|---|---|
| tier1-dpdk-echo | arp_resolution | ✅ pass | 2.412s | |
| tier1-dpdk-echo | udp_send_receive | ✅ pass | 1.752s | |
| tier1-dpdk-echo | echo_roundtrip | ✅ pass | 1.551s | |
| tier1-dpdk-echo | payload_integrity | ✅ pass | 0.744s | |
| tier1-dpdk-echo | jumbo_diagnostics | ✅ pass | 0.025s | |
| tier1-dpdk-echo | jumbo_echo_8000 | ✅ pass | 1.746s | |
| tier1-tcp-echo-async | execution | 0.000s | ENI bind failed on receiver instance | |
| tier1-tcp-echo | bidir_echo | ✅ pass | 0.736s | |
| tier1-tcp-echo | bidir_multi | ✅ pass | 0.771s | |
| tier2-kernel-interop | arp_resolution | ✅ pass | 0.064s | |
| tier2-kernel-interop | udp_send_receive | ✅ pass | 1.079s | |
| tier2-kernel-interop | echo_roundtrip | ✅ pass | 0.871s | |
| tier2-kernel-interop | payload_integrity | ✅ pass | 0.068s | |
| tier2-tcp-echo | execution | 0.000s | ENI bind failed on receiver instance | |
| tier3-iperf-interop | our_app_sends | ✅ pass | 2.568s |
⚠️ Infra flakes (2) — SSM/ENI setup, not code (task #10)
- tier1-tcp-echo-async / execution — ENI bind failed on receiver instance
- tier2-tcp-echo / execution — ENI bind failed on receiver instance
❌ Integration Tests Failed (Run 33505065263)Branch: Test Results
Application Logs (last 20 lines)sender-echo-server.log sender-test-client.log sender-test-client-iperf.log Full Application Logs (last 200 lines each)sender-echo-server.logsender-test-client.logsender-test-client-iperf.log
|
|
Wire the existing QUIC Tier 1 harness into the shared integration orchestrator, add QUIC integration and performance CI, and define evidence-based gates for graduating TCP and QUIC from non-blocking to blocking CI. Spec includes: - requirements.md: 10 functional requirements, 4 review rounds - design.md: orchestrator pseudocode, CI workflow structure, gate runbook - tasks.md: 19 implementation tasks split into 2 ROADMAP items Review: 4 Claude opus-5 rounds (SHIPPABLE, 0 CRITICAL / 0 HIGH remaining) EARS: clean
gspivey
force-pushed
the
feat/quic-tcp-ec2-integration-spec
branch
from
October 11, 2026 02:29
276be5a to
8394723
Compare
Synthetic Performance Results — Graviton (run)Commit: ✅ synthetic UDP socket bound to 10.0.0.1:9000 (MAC: 02:00:00:00:00:01) Synthetic UDP Performance ResultsMeasures framework overhead: sync IPv4 Baseline
IPv6
IPv6 vs IPv4 Comparison (sync path)
IPv4 avg sync/async ratio: 0.9x, worst: 1.1x | IPv6 vs IPv4 worst ratio: 1.32x (OK)
|
TCP Synthetic Performance Results (run)Commit: TCP Synthetic Performance ResultsMeasures TCP engine framework overhead using a mock PacketBackend.
conn_est=1223ns, throughput=6181.2Mbps, tick=481ns |
Synthetic Performance Results (run)Commit: ✅ synthetic UDP socket bound to 10.0.0.1:9000 (MAC: 02:00:00:00:00:01) Synthetic UDP Performance ResultsMeasures framework overhead: sync IPv4 Baseline
IPv6
IPv6 vs IPv4 Comparison (sync path)
IPv4 avg sync/async ratio: 0.9x, worst: 1.1x | IPv6 vs IPv4 worst ratio: 1.32x (OK)
|
[CI] Stage: DeployInfrastructure ready.
|
Integration Tests — ❌ 3 real failures
❌ Real failures (3) — code/test issues
detailsdetails |
Integration Test Failure — Graviton (Run 38105318120)Branch: receiver-user-data.log (7305 bytes, last 80 lines)sender-user-data.log (7305 bytes, last 80 lines)receiver-console-output.log (45235 bytes, last 80 lines)sender-console-output.log (45697 bytes, last 80 lines) |
❌ Integration Tests Failed — Graviton (run)Branch: Test Results
Application Logs (last 20 lines)receiver-echo-server.log sender-echo-server.log sender-test-client.log |
[CI] Stage: DeployInfrastructure ready.
|
|
gspivey
added a commit
that referenced
this pull request
Oct 13, 2026
## ROADMAP items 39–42 Spec for the 3 root-caused TCP integration-test failures that were found after PR #122 merged. ## Root causes (confirmed against actual code) **Bug 1 — RFC 6191 TIME_WAIT (engine.rs lines ~1557–1578)** `handle_time_wait` only handles FIN retransmits; any other packet including SYN is silently dropped. Fix: add RST handling (remove TCB), SYN > rcv_nxt handling (new incarnation), SYN ≤ rcv_nxt (re-ACK to trigger RST→SYN-retransmit cycle). **Note:** ISN generator uses per-process random secret (confirmed in isn.rs), so RFC 6191 ISN comparison alone is ~50/50 — full RST recovery path is required. **Bug 2 — Asymmetric routing (Phase C kernel_to_dpdk_echo)** Kernel client sends from eth0 (management IP); DPDK receiver's SYN-ACK returns to eth1 subnet; rp_filter=1 drops it. Fix: add `--bind-ip` to tcp-test-client, pass ETH1_IP in Phase C harness. **Bug 3 — vfio-pci bind race (Graviton-specific)** EAL init fires before /dev/vfio/noiommu-* node appears after dpdk-devbind.py. Fix: poll driver symlink + device node with 60s timeout before launching any DPDK binary. ## SOP exception Item 39 (known-failures gate) is a prerequisite. Once merged, items 40–42 each remove one entry from `ci/known-failures.txt`. CI is green when the target test passes AND all remaining listed tests fail as expected. ## Review history Spec designed with Claude opus-5.5 + actual code inspection (isn.rs, engine.rs). Key finding: the ISN coin-flip issue changes the engine fix from a one-liner to the full RFC 1122 §4.2.2.13 recovery sequence. <!-- no-visual-delta --> **Why no screenshot:** spec + ROADMAP only.
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
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.
Summary
Adds spec for QUIC + TCP EC2 integration and performance CI gaps.
What's Missing Today (and What This Spec Fixes)
run_quic_tier1function--tier quicfilter + cert transfer via SSMquic-integration-tests.ymlworkflowcontinue-on-erroruntil 10 clean runsquic-perf-tests.ymlworkflowrun-perf-tests.sh --configs quic-native-dpdk-niccontinue-on-errorgate: no defined removal procedureSpec Contents
run_quic_tier1,quic_transfer_cert,quic_verify_build; CI workflow YAML structure; gate counting runbookReview History
ROADMAP Items Added
Key Design Decisions
quic-only: not part of run-all, so primaryintegration-testsjob is unchangedconcurrency.group: integration-tests(can't useneeds:across workflow files)run-quic-integration-tests.shsed-escaping approach (which is vulnerable to%in cert content)ssm_run_get_outputarg order:(instance_id, timeout, cmd)— deliberately reordered fromrun-quic-perf.sh:175's(instance_id, cmd, timeout)to match the orchestrator convention