deploy.yml ran `compose up -d` before the health check and never reverted
on failure. A build that passes tests, returns 200 on /health with the
right git_sha, but has a live logic bug (exactly the class of bug the
correlator instrumentation exists to catch) would stay live indefinitely
- notify-failure would even claim production was "still running the
previous build", which is false in that scenario.
Deploy step now reads /opt/drb/.last_good_tag (written only after a prior
deploy's own health check confirmed its SHA) to capture the previously-
verified tag before switching, and emits it as a step output. Health
check is unchanged in shape (bounded 20x5s retry, still requires the
polled git_sha to match) but now persists the new SHA as the rollback
target only once confirmed live. A new Rollback step runs on any failure
above, re-deploys the previous tag, and re-verifies via the same git_sha
check rather than trusting mere liveness - then fails the job loudly
either way, since the push itself was still bad. notify-failure now
reports what actually happened (rollback succeeded/failed/skipped and to
which SHA) instead of the old unconditional claim.
This unblocks #62 decision 9: autonomous pushes to incident_correlator.py,
llm_correlator.py, intelligence.py and routers/upload.py were frozen until
this rollback path landed.
Refs #65, #62, #60, #57.
The build-stamp health check added in 8fbfe7d worked on its first run, and
what it caught was not a stale container -- it was a race. Runs 544 and 545
overlapped; both deployed :latest, 545's images won, and 544's health check
correctly reported that the build serving traffic was not the one it had just
deployed.
That is a real hazard, not a false positive: with :latest, two pushes landing
close together means whichever finishes last silently wins for BOTH, and
neither run's log tells you which code is actually live. Pushes land close
together constantly here.
docker-compose.yml already resolved images as ${TAG:-latest}, so the fix is to
export TAG=<commit sha> for the deploy. Each run now pulls and starts exactly
the images it built, rollback becomes "deploy a different tag", and the health
check's assertion becomes meaningful rather than order-dependent. A manual
`docker compose up -d` on the VM with no TAG set still falls back to :latest,
which is the intended escape hatch.
Also replaces the health check's single `sleep 20` with a poll of up to 100s
that stops as soon as the expected SHA appears. A fixed sleep is either too
short -- flaky red runs -- or wastes time on every deploy, and a check that
cries wolf gets ignored, which is exactly the failure this job exists to stop.
Refs logan/server-26#21
Two unrelated-looking problems with the same shape: a dangerous state that
looked fine from the outside.
DEPLOY (server-26#21). The Deploy job failed on fifteen consecutive pushes
between 2026-08-18 and 08-20 and nobody noticed for two days, because the
build job was green and a red run is only visible to someone who opens Gitea.
Production served 08-18 code the whole time -- including the entire frontend
redesign, chunks 2 through 8. Three changes:
* The health check now asserts WHICH build answered, not just that something
did. CI bakes the commit into the image (Dockerfile ARG/ENV GIT_SHA) and
/health reports it, so a deploy that "succeeds" while the previous
container keeps running now fails. Liveness alone could never have caught
this.
* The image pull retries once after a prune. The actual failure was
containerd unable to extract a layer -- "failed to Lchown ... no such file
or directory" -- a corrupted entry in the snapshot store, which a prune
clears. A second failure after pruning is a real problem (check the VM's
disk) and still stops the deploy.
* A notify-failure job POSTs to DEPLOY_ALERT_WEBHOOK when anything in the
workflow fails. Unset means skip quietly, not fail.
CORS (server-26#20). allow_origins=["*"] with allow_credentials=True is not
the permissive-but-harmless setting it reads as. Starlette does not reject the
pair -- it reflects the caller's Origin back and still sends
Access-Control-Allow-Credentials: true, so the effective policy is "any
origin, WITH credentials", the opposite of what a wildcard normally means.
Rather than trust every deployment to remember CORS_ORIGINS, the pair is now
unrepresentable: a wildcard forces allow_credentials off and logs an ERROR
naming the variable to set. Correctly configured deployments that name their
origins are unaffected and keep credentialed requests.
Severity honestly: low today. c2-core is bearer-auth, and browsers do not
attach bearer tokens cross-origin the way they attach cookies. This is a
misconfiguration waiting for the day something starts trusting a cookie.
Also adds firebase_admin.auth.UserRecord and the list/update/create/delete_user
names to the conftest stub. routers/users.py annotates with UserRecord at
import time, so without it importing app.main failed at collection -- which is
why nothing had ever tested anything wired at app level, CORS included.
Tests: 5 new in test_cors_policy.py, covering the pure policy function, the
middleware actually mounted on the app (so re-hardcoding allow_credentials=True
fails here), and the presence of the build stamp.
Closeslogan/server-26#20Closeslogan/server-26#21
Shallow clones were never a Gitea packing bug. An intruder had set
uploadpack.packObjectsHook in Gitea's HOME gitconfig, pointing at a
non-executable dropper, so every upload-pack died mid-pack. That hook is
gone and --depth=1 clones are verified working, so fetch-depth: 0 buys
nothing but slower CI. See INCIDENT-2026-08-11.md.
infra/Caddyfile was dead: ansible templates Caddyfile.j2 to
/etc/caddy/Caddyfile, and nothing ever deployed the static copy.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
actions/checkout defaults to depth=1, and Gitea aborted generating that pack
with a bad pack header protocol error on all three retries, failing the build
before any image was pushed. A full fetch avoids the shallow-pack path.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Edge nodes are deployed to arbitrary locations by arbitrary people, so the
broker has to be reachable from the internet and secured on its own merits
rather than by a VPN.
Three defects made that impossible. The broker only had a plaintext 1883
listener; every node shared one drb-node password; and the ACL pattern used
%c, the client-supplied client id, so any holder of that shared password
could set client_id to another node and take over its namespace. The comment
claiming this cryptographically prevented cross-node access was wrong and is
gone.
Authentication now uses mosquitto 2.x's built-in dynamic-security plugin on
the stock eclipse-mosquitto image. c2-core administers it over the control
topic, creating each node's client on approval with username=<node_id> and
password=<its node_keys api_key>, attached to a role whose ACL is nodes/%u/#
against the authenticated username. One credential, one revocation point.
An HTTP-callback plugin was implemented first and rejected: that project is
archived upstream, which is not an acceptable dependency on an
internet-facing broker.
Because dynsec state is a second source of truth alongside Firestore,
approve/reissue/delete now write to the broker first and surface a 502
rather than drifting, and c2-core reconciles every approved node into dynsec
on startup.
Adds node self-enrollment (POST /nodes/enroll, GET /nodes/{id}/credentials)
so a new node can obtain its key over HTTPS without an operator handling
secrets by hand. Enrolling an already-approved node_id is refused on the
fleet token alone — otherwise a leaked token plus a guessable id would let
an attacker steal a live node's key before the real node asked for it.
Pickup secrets are stored hashed and returned once, and the endpoint is rate
limited per source IP.
Infrastructure: an 8883 TLS listener fed by Caddy's certificate via a
systemd path unit, a firewall rule for it, and Caddy now 404s /internal/*
so the api vhost cannot proxy internal routes.
Also fixes CORS, which allowed https://app.<domain> while the frontend is
served on the bare domain — every call from the portal would have failed —
and widens the vault gitignore to a glob, since ansible-vault leaves
backup siblings that the exact-name rule left committable.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
- Terraform: e2-micro VM (us-east1-b, free tier), static IP, SSH/web
firewall rules, IAM bindings for Firestore + GCS; imports existing
drb-calls bucket and c2-server Firestore database into state
- Gitea CI: build c2-core, discord-bot, frontend images and push to
git.vpn.cusano.net registry; SSH deploy pulls pre-built images (no
build on VM)
- Ansible: first-time setup only — git clone, env files from vault,
Caddyfile, docker login + compose pull + up; no rsync or on-VM builds
- docker-compose: add image: ${REGISTRY}/name:latest alongside build:
so local dev and CI registry both work
- gitignore: add Terraform state, lock, tfvars, ansible secrets
Logan CusanoandClaude Sonnet 5