From a1e96a1dc4e7ed8a07fcebbcdb3337e31604a779 Mon Sep 17 00:00:00 2001 From: Jinwoo-H Date: Sat, 5 Sep 2026 21:04:21 -0400 Subject: [PATCH] docs(relay): record Roll 2 phase 0/1 (code merge, image, director deploy) --- cloud/docs/relay-reconnect-2026-09-findings.md | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/cloud/docs/relay-reconnect-2026-09-findings.md b/cloud/docs/relay-reconnect-2026-09-findings.md index 426a120c251..580a4da84d8 100644 --- a/cloud/docs/relay-reconnect-2026-09-findings.md +++ b/cloud/docs/relay-reconnect-2026-09-findings.md @@ -989,3 +989,14 @@ Owner: "sure, feel free to drive these." Sequence chosen: Roll 1 first (highest | Monitor dry-run #50 | Dispatched 21:54Z at gen 146 on main `51eed5a1bc`, run 33994385666. **Green** 22:10Z, 16/16 samples. Main had moved to `d7767fb196`; trusted paths identical to `a3c1d32995`. Chain dispatched the c29 `canary-apply` (run 33995164002, protocol 0) 12 s after green. | | c29 canary (run 33995164002, `canary-apply`) | **Success** 22:27Z. Isolate → migration-only at **gen 147**, verifier passed on the old image (1 199 assignments), Terraform applied same-cap template `…20260905221622`, new incarnation on `519f4914` at protocol 0, verifier passed at migration-only, activate → **gen 148**, c29 general, verifier passed (1 199 assignments carried). No `container die` fleet-wide 22:11Z–22:30Z. | | **Roll 1 complete** | Image census 22:30Z from MIG templates: c8–c10, c13–c16, c19–c29 on `519f4914` (18 cells); c7 on `85bf6799` (the earlier rehearsal image, carries the same fix); existing-only c1–c6, c11, c12 and migration-only c17, c18 untouched by design. No serving cell remains on `5aedbca5`. Selector gen 148, membership unchanged from the start of the roll. Zero relay container exits fleet-wide across the roll (01:14Z–22:30Z). Gates used: #19–#50; freezes were all monitor-side (provenance, freshness, flat Asia latency bar, one Cloud Monitoring collector failure), none a fleet health finding. Roll 2 (fresh image with #18722 + #18720) is the next data-plane step and waits on the owner's private-IP window decision. | + +## Roll 2 (image `4916ed67`, 2026-09-06) + +| Step | Result | Evidence | +|---|---|---| +| Docs split | #18958 merged `3bb038a185` (findings, checklist, roadmap, Roll 2 plan). | | +| Code PR | #18959 merged `61b09b7a02` (rebase of #18565 onto main; desktop rotation change dropped since #18719 shipped a proportional version). Two Opus review rounds: round 1 caught the mobile fail-fast rejecting on any socket close (one AP flap would book the 60 s cooldown) → 2 s grace, re-armed once on `handshaking`; round 2 caught a removed jitter assertion that let a one-sided jitter pass → exact pin on the top of the band. Control lease 55 min → 6 h ± 30 min. | | +| Image publish | run 34002233801 → `sha256:4916ed676d8389f694a648e750f1112d9002d68c84a1e0c7af828d5af129de62`; mirrored to staging (run 34002326150). | | +| Staging cell smoke | **Dropped.** Staging C4 is pinned to the Asia launch digest by `relay-staging-c4-refresh-workflow.test.mjs` (with production c27–c29 tfvars and the C4 recovery workflow) and the only C4 image-refresh path pins its accepted predecessor to an older digest. Re-pinning all of it for a smoke widens into the Asia launch machinery; #18969 closed. Roll 2 follows the Roll 1 path: director first, c7 as the rehearsal cell. | | +| Director deploy | run 34002673626 **success** 01:02Z: serving `orca-cloud-relay-00575-leq` on `4916ed67`, `00574-wag` (same image) tagged `selector-rollback`, `00569-ret` (`519f4914`) still deployable. Baseline before: 1 director Postgres retry in the prior hour, 0 `container die`. | | +| c7 `verify` (read-only) | run 34002885408 dispatched 01:03Z, target `4916ed67`, rollback `85bf6799`, protocol 1, gen 148. | |