release v1.2.2: edge HTTP/3 stall fix + db-size dashboard threshold #121

Merged

developer merged 4 commits from development into master 2026-06-22 19:50:58 +00:00

Conversation 1 Commits 4 Files Changed 4

+142 -2

developer commented 2026-06-22 19:42:37 +00:00

Owner

Copy Link

Copy Source

Release v1.2.2 — promote development → master

Two fixes, both config/dashboard only — no schema migration, so the rollout is
image/config-only, rollback-safe, and needs no maintenance window.

Included

• #119 fix(edge): suppress dead HTTP/3 advert (Alt-Svc: clear). The edge
  advertised HTTP/3 but UDP/443 is never exposed (compose maps only TCP, ufw opens
  443/tcp), so clients cached the 30-day Alt-Svc: h3 and stalled on a dead QUIC path
  before falling back to h2 — the Telegram Mini App intermittently hung on load.
  Emitting Alt-Svc: clear site-wide pins clients to h2/h1 and actively flushes the
  poisoned cache. Confirmed live on the test contour (host caddy). Background +
  alternatives: docs/EDGE_HTTP3.md.
• #120 fix(grafana): real byte thresholds for the Database size stat. The panel had
  no thresholds, so Grafana's default (red at >=80) painted a healthy ~9 MiB DB
  permanently red. Now green / yellow 8 GiB / red 16 GiB, sized to the 40 GiB prod disk.

Scope: affects the main host only (caddy + grafana); the bot host is untouched.

After merge (manual, owner)

1. Tag v1.2.2 on master HEAD.
2. Prod deploy: prod-deploy.yaml, workflow_dispatch, confirm=deploy from master.
   The h3 fix takes effect on prod only after this deploy (and then flushes already-cached
   h3 entries on real clients via Alt-Svc: clear).

# Release v1.2.2 — promote `development` → `master` Two fixes, both **config/dashboard only — no schema migration**, so the rollout is image/config-only, rollback-safe, and needs no maintenance window. ## Included - **#119 `fix(edge)`: suppress dead HTTP/3 advert (`Alt-Svc: clear`).** The edge advertised HTTP/3 but UDP/443 is never exposed (compose maps only TCP, ufw opens 443/tcp), so clients cached the 30-day `Alt-Svc: h3` and stalled on a dead QUIC path before falling back to h2 — the **Telegram Mini App intermittently hung on load**. Emitting `Alt-Svc: clear` site-wide pins clients to h2/h1 and actively flushes the poisoned cache. Confirmed live on the test contour (host caddy). Background + alternatives: `docs/EDGE_HTTP3.md`. - **#120 `fix(grafana)`: real byte thresholds for the Database size stat.** The panel had no thresholds, so Grafana's default (red at `>=80`) painted a healthy ~9 MiB DB permanently red. Now green / yellow 8 GiB / red 16 GiB, sized to the 40 GiB prod disk. Scope: affects the **main host** only (caddy + grafana); the bot host is untouched. ## After merge (manual, owner) 1. Tag `v1.2.2` on `master` HEAD. 2. Prod deploy: `prod-deploy.yaml`, `workflow_dispatch`, `confirm=deploy` from `master`. The h3 fix takes effect on prod only after this deploy (and then flushes already-cached `h3` entries on real clients via `Alt-Svc: clear`).

developer added 4 commits 2026-06-22 19:42:37 +00:00

fix(edge): suppress dead HTTP/3 advert with Alt-Svc: clear ... 9253b1bdca

Caddy enables HTTP/3 by default on any TLS listener and emits
Alt-Svc: h3=":443"; ma=2592000, but UDP/443 is never reachable: the prod
compose maps only "443:443" (TCP) and ufw opens 443/tcp (test contour: the
host caddy publishes only :443/tcp). A client that cached the 30-day advert
tries QUIC first on later opens, gets no response, and waits for the QUIC
attempt to time out before falling back to h2 -- which surfaced as the
Telegram Mini App intermittently hanging on load (a barely-noticeable pause
up to a blank window). The h2/TCP serving path itself is healthy (~10ms TTFB).

Emit Alt-Svc: clear site-wide at the contour caddy so clients actively drop
any cached alternative and stay on h2/h1. This caddy terminates TLS in prod
(the fix target); in the test contour it serves plain :80 and the host caddy
re-stamps its own Alt-Svc, so the live test fix lives in the host caddy. Add
docs/EDGE_HTTP3.md (symptom, diagnosis method, verify, and option B -- serving
h3 for real -- if it recurs) and link it from ARCHITECTURE.md.

fix(grafana): give "Database size" stat real byte thresholds ... 46d569720c

The "Database size" stat had no thresholds, so Grafana applied its stat
default (green base, red at >=80). The query is pg_database_size_bytes, so a
healthy ~9 MiB database (9.4M >> 80) rendered permanently RED on the
Scrabble - Resources dashboard (test + prod), reading as a false alert; the
neighbouring percentunit cache-hit stat stayed green only because its 0..1
values fall under 80.

Add absolute byte thresholds sized to the 40 GiB prod disk (4.6 GiB used,
observability bounded -- Tempo <=1 GiB, Prometheus 7d -- so the DB is the
only unbounded grower): green up to 8 GiB, yellow at 8 GiB (~20% of disk),
red at 16 GiB (~40%), an early warning with ample runway before the disk
tightens, not a panic line. Cosmetic panel coloring only; there are no
Grafana alert rules provisioned.

Merge pull request 'fix(edge): suppress dead HTTP/3 advert with Alt-Svc: clear' (#119) from feature/edge-suppress-http3-altsvc into development 6cb88b28c4

Merge pull request 'fix(grafana): real byte thresholds for the Database size stat' (#120) from feature/grafana-db-size-thresholds into development ba57687430

owner approved these changes 2026-06-22 19:45:08 +00:00

developer merged commit dc946a1faf into master 2026-06-22 19:50:58 +00:00

Sign in to join this conversation.

Reviewers

No Reviewers

owner

Labels

No items

No Label

Milestone

No items

No Milestone

Projects

Clear projects

No project

Assignees

Clear assignees

owner (Ilia) developer

No Assignees

2 Participants

Notifications

Subscribe

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: developer/scrabble-game#121