bao/nexus
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: harden Nexus runtime health and rollback

Browse Source
This commit is contained in:
DevOps
2026-07-10 00:37:24 +02:00
parent dbda764190
commit 706ff82ccd
7 changed files with 106 additions and 68 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.env.template
+2 -2
View File
@@ -19,7 +19,7 @@ JWT_AUDIENCE=nexus-web
BOOTSTRAP_OWNER_EMAIL=***
# ── OpenClaw Integration ────────────────────────────────
# Base URL of the OpenClaw gateway (host.docker.internal from inside container)
OPENCLAW_BASE_URL=http://host.docker.internal:18789
# Internal Docker-DNS URL of the OpenClaw gateway
OPENCLAW_BASE_URL=http://openclaw-gateway-bao:18789
OPENCLAW_GATEWAY_TOKEN=***
OPENCLAW_GATEWAY_PASSWORD=***
.gitea/scripts/deploy-nexus.sh
+10 -4
View File
@@ -60,7 +60,7 @@ JWT_KEY=${ENV_JWT_KEY}
JWT_ISSUER=nexus
JWT_AUDIENCE=nexus-web
BOOTSTRAP_OWNER_EMAIL=vmbao62@hotmail.de
OPENCLAW_BASE_URL=http://host.docker.internal:18789
OPENCLAW_BASE_URL=http://openclaw-gateway-bao:18789
OPENCLAW_GATEWAY_TOKEN=${ENV_OPENCLAW_TOKEN:-}
OPENCLAW_GATEWAY_PASSWORD=
NEXUS_VERSION=${VERSION}
@@ -140,9 +140,15 @@ echo "Checking live health"
retry=0
while [ "$retry" -lt 6 ]; do
retry=$((retry + 1))
if curl -fsS --max-time 10 "$BASE_URL/health" >/dev/null; then
echo "Health check passed"
break
health_body="$(curl -fsS --max-time 10 "$BASE_URL/health" 2>/dev/null || true)"
case "$health_body" in
'{"status":"Healthy"'*)
echo "Health check passed"
break
;;
esac
if [ -n "$health_body" ]; then
echo "Health endpoint is reachable but not healthy: $health_body" >&2
fi
if [ "$retry" -eq 6 ]; then
echo "Health check failed" >&2
.gitea/workflows/rollback.yaml
+50 -23
View File
@@ -112,9 +112,11 @@ jobs:
JWT_ISSUER=nexus
JWT_AUDIENCE=nexus-web
BOOTSTRAP_OWNER_EMAIL=vmbao62@hotmail.de
OPENCLAW_BASE_URL=http://host.docker.internal:18789
OPENCLAW_BASE_URL=http://openclaw-gateway-bao:18789
OPENCLAW_GATEWAY_TOKEN=${ENV_OPENCLAW_TOKEN}
OPENCLAW_GATEWAY_PASSWORD=
NEXUS_VERSION=$(tr -d '[:space:]' < VERSION)
NEXUS_GIT_SHA=$(git rev-parse HEAD)
EOF
chmod 600 "${ENV_TMPFILE}"
@@ -127,18 +129,38 @@ jobs:
run: |
set -euo pipefail
docker run --rm \
-v "${{ gitea.workspace }}:/src:ro" \
git archive --format=tar HEAD | docker run --rm -i \
-v "${DEPLOY_PATH}:/dest" \
alpine:latest \
sh -c "
cd /src && \
find . -mindepth 1 -maxdepth 1 \
! -name .git \
-exec cp -r {} /dest/ \; && \
DEST_OWNER=\$(stat -c '%u:%g' /dest) && \
chown -R \"\$DEST_OWNER\" /dest
"
sh -c '
set -eu
dest_owner="$(stat -c "%u:%g" /dest)"
mkdir -p /src-snapshot
tar -xf - -C /src-snapshot
is_protected_path() {
case "$1" in
./.git|./.env|./.env.*|./data|./logs|./backups|./tmp|./uploads|./storage)
return 0
;;
*)
return 1
;;
esac
}
cd /dest
find . -mindepth 1 -maxdepth 1 | while IFS= read -r path; do
if ! is_protected_path "$path"; then rm -rf "$path"; fi
done
cd /src-snapshot
find . -mindepth 1 -maxdepth 1 | while IFS= read -r path; do
if ! is_protected_path "$path"; then cp -a "$path" /dest/; fi
done
chown -R "$dest_owner" /dest
'
echo "✅ Rollback code (${{ inputs.target_tag }}) synced to ${DEPLOY_PATH}"
@@ -151,16 +173,18 @@ jobs:
docker run --rm \
-v "${DEPLOY_PATH}:/workspace/nexus" \
-v "/tmp:/tmp-host:ro" \
-v /var/run/docker.sock:/var/run/docker.sock \
-w /workspace/nexus \
-i \
docker:cli \
sh -c "
set -e
echo '🔙 Rolling back to ${{ inputs.target_tag }}'
docker compose --env-file /tmp-host/$(basename "${ENV_TMPFILE}") build --no-cache
docker compose --env-file /tmp-host/$(basename "${ENV_TMPFILE}") up -d --wait --force-recreate
"
sh -c '
set -eu
umask 077
cat > /tmp/nexus-rollback-env
trap '\''rm -f /tmp/nexus-rollback-env'\'' EXIT INT TERM
docker compose --env-file /tmp/nexus-rollback-env build --no-cache
docker compose --env-file /tmp/nexus-rollback-env up -d --wait --force-recreate
' < "${ENV_TMPFILE}"
echo "✅ Rollback redeploy completed"
@@ -186,11 +210,14 @@ jobs:
WAIT=1
while [ $RETRY -lt $MAX ]; do
RETRY=$((RETRY + 1))
if curl -sf --max-time 10 https://nexus.noveria.net/health; then
echo ""
echo "✅ Health check passed (attempt $RETRY/$MAX)"
exit 0
fi
HEALTH_BODY=$(curl -sf --max-time 10 https://nexus.noveria.net/health || true)
case "$HEALTH_BODY" in
'{"status":"Healthy"'*)
echo "✅ Health check passed (attempt $RETRY/$MAX)"
exit 0
;;
esac
[ -n "$HEALTH_BODY" ] && echo "⚠️ Health endpoint is degraded: $HEALTH_BODY"
echo "⏳ Attempt $RETRY/$MAX failed, waiting ${WAIT}s..."
sleep $WAIT
NEXT=$((WAIT + RETRY))
compose.yaml
+12 -3
View File
@@ -2,6 +2,15 @@ name: nexus
services:
postgres:
image: postgres:17-alpine
# WAL-Archivierung bleibt deaktiviert, bis ein verwaltetes Off-Server-Ziel
# mit Retention und Restore-Test existiert. Ein lokales Endlosarchiv ist
# kein Backup und kann bei Fehlern pg_wal ungebremst wachsen lassen.
command:
- postgres
- -c
- archive_mode=off
- -c
- archive_command=
restart: always
deploy:
resources:
@@ -50,7 +59,7 @@ services:
Jwt__Audience: ${JWT_AUDIENCE:-nexus-web}
Bootstrap__OwnerEmail: ${BOOTSTRAP_OWNER_EMAIL:?Set BOOTSTRAP_OWNER_EMAIL in .env}
# Initial owner password is generated once at first seed and then lives only in the DB.
Integrations__OpenClaw__BaseUrl: ${OPENCLAW_BASE_URL:-http://host.docker.internal:18789}
Integrations__OpenClaw__BaseUrl: ${OPENCLAW_BASE_URL:-http://openclaw-gateway-bao:18789}
Integrations__OpenClaw__Token: ${OPENCLAW_GATEWAY_TOKEN:-}
Integrations__OpenClaw__Password: ${OPENCLAW_GATEWAY_PASSWORD:-}
Admin__ResetToken: ${Admin__ResetToken:-}
@@ -59,7 +68,7 @@ services:
- host.docker.internal:host-gateway
depends_on:
postgres:
condition: service_started
condition: service_healthy
restart: true
healthcheck:
test: ["CMD-SHELL", "wget --no-verbose --tries=1 --spider http://localhost:8080/health/live || exit 1"]
@@ -106,7 +115,7 @@ services:
- "127.0.0.1:18880:80"
depends_on:
api:
condition: service_started
condition: service_healthy
restart: true
healthcheck:
test: ["CMD-SHELL", "curl -f http://localhost:80/ || exit 1"]
docs/architecture-board-first-orchestration.md
+16 -21
View File
@@ -68,8 +68,8 @@ Ansatz. Das Backend fungiert bereits als sichere Schicht zwischen allen Akteuren
│ │
│ ALLE Gateway-Calls → Authorization: Bearer <Gateway-Password> │
└──────────────┬──────────────────────────────┬────────────────────┘
│ host.docker.internal:18789 │
│ (Gateway loopback/lan) │
│ openclaw-gateway-bao:18789 │
│ (internes Docker-DNS) │
▼ │
┌──────────────────────────────┐ │
│ OpenClaw Gateway Container │ │
@@ -199,7 +199,7 @@ Ebene 4: X-Agent-Id Header (Agent-Identität für Task-State-Enforcement)
```
POST /api/v1/operations/snapshot
→ DashboardService → OpenClawGatewayClient.InvokeToolAsync()
→ POST http://host.docker.internal:18789/tools/invoke
→ POST http://openclaw-gateway-bao:18789/tools/invoke
Authorization: Bearer <Gateway-Password>
```
@@ -220,7 +220,7 @@ POST /api/v1/operations/snapshot
### 5.2 Docker-Netzwerk & Gateway-Bind
**Aktuelles Problem:**
**Aktueller Stand (2026-07-09):**
```
compose.yaml:
api:
@@ -228,32 +228,27 @@ compose.yaml:
- host.docker.internal:host-gateway
networks:
- nexus
- openclaw_default ← API-Container ist im Gateway-Netzwerk
- openclaw_default
Gateway-Konfiguration:
gateway.bind: "loopback" ← Bindet nur 127.0.0.1 IM GATEWAY-CONTAINER
gateway.bind: "lan"
Nexus-Konfiguration:
OPENCLAW_BASE_URL=http://openclaw-gateway-bao:18789
```
**Ergebnis:**
- `host.docker.internal:18789` funktioniert, weil `extra_hosts` auf den Docker-Host zeigt
- ABER: Docker-Port-Forward (wenn vorhanden) sendet an Container-IP, nicht loopback
- Die `openclaw_default` Netzwerk-Mitgliedschaft des API-Containers wird NICHT genutzt
- Nexus erreicht das Gateway direkt über Docker-DNS im gemeinsamen `openclaw_default`-Netz.
- Der Umweg über einen nicht veröffentlichten Host-Port entfällt.
- Der produktive Aggregat-Healthcheck prüft neben PostgreSQL auch die Runtime-Verbindung.
**Empfehlung (siehe gateway-api-research.md, Abschnitt 6):**
```json5
// openclaw.json
{
gateway: {
bind: "lan" // war "loopback"
}
}
```
Der frühere Pfad `host.docker.internal:18789` war auf dem VPS nicht erreichbar und ist obsolet.
Alternativ: API-Container über Gateway-Container-Namen ansprechen:
Produktive Einstellung:
```yaml
Integrations__OpenClaw__BaseUrl: http://openclaw_gateway:18789
Integrations__OpenClaw__BaseUrl: http://openclaw-gateway-bao:18789
```
(Vorausgesetzt der Gateway-Container heißt `openclaw_gateway` und ist im `openclaw_default` Netzwerk)
Beide Container müssen Mitglied im `openclaw_default`-Netzwerk sein.
### 5.3 MCP-artige Integration: Bewertung
docs/gateway-api-research.md
+13 -13
View File
@@ -290,30 +290,30 @@ The Nexus compose.yaml already includes the full integration infrastructure:
```yaml
api:
extra_hosts:
- host.docker.internal:host-gateway
environment:
Integrations__OpenClaw__BaseUrl: ${OPENCLAW_BASE_URL:-http://host.docker.internal:18789}
Integrations__OpenClaw__BaseUrl: ${OPENCLAW_BASE_URL:-http://openclaw-gateway-bao:18789}
Integrations__OpenClaw__Token: ${OPENCLAW_GATEWAY_TOKEN:-}
Integrations__OpenClaw__Password: ${OPENCLAW_GATEWAY_PASSWORD:-}
networks:
- nexus
- openclaw_default
```
The API container:
- Uses `host.docker.internal:18789` to reach the Gateway via the Docker host
- Has `extra_hosts` configured for `host.docker.internal`
- Uses Docker DNS (`openclaw-gateway-bao:18789`) in the shared `openclaw_default` network
- Does not depend on a published host port for the Gateway
- Reads token/password from `.env` via `OPENCLAW_GATEWAY_PASSWORD`
### Known Issue: Gateway Bind = loopback
### Resolved Routing Issue (2026-07-09)
The Gateway binds to `127.0.0.1` (`gateway.bind: "loopback"`). This means it only listens inside the gateway container's loopback interface.
The old `host.docker.internal:18789` route was unreachable because no usable host port was published. The Gateway is now reached directly by its container DNS name.
| Scenario | Works? | Why |
|----------|--------|-----|
| Gateway with `--network host` | ✅ Yes | Process sees host's 127.0.0.1 directly |
| Gateway with `-p 18789:18789` + loopback bind | ❌ No | Port forward sends to container IP, not loopback |
| Gateway with `-p 18789:18789` + lan bind | ✅ Yes | Listens on all interfaces including container IP |
| `host.docker.internal:18789` | ❌ No | No reachable host listener on the VPS |
| `openclaw-gateway-bao:18789` in `openclaw_default` | ✅ Yes | Direct container-to-container routing via Docker DNS |
**Fix**: Change `gateway.bind` from `"loopback"` to `"lan"` (binds `0.0.0.0`):
The Gateway must listen on its container interface (`gateway.bind: "lan"`):
```json5
{
@@ -325,8 +325,8 @@ The Gateway binds to `127.0.0.1` (`gateway.bind: "loopback"`). This means it onl
**Test command (from Nexus API container):**
```bash
curl -s http://host.docker.internal:18789/health
# Expected: 200 if gateway bind is lan/container IP is reachable
curl -s http://openclaw-gateway-bao:18789/
# Expected: HTTP 200 from inside nexus-api-1
```
### Required .env Vars for Nexus
phases/runtime-routing.md
+3 -2
View File
@@ -1,6 +1,6 @@
# Runtime und Routing
> Letzte Aktualisierung: 2026-06-16
> Letzte Aktualisierung: 2026-07-09
## Aktive Modelle (7 von 8 konfiguriert)
@@ -23,6 +23,7 @@
- Einzige aktive Integration: `OpenClawRuntime` über `IAgentRuntime`
- Model-Routing läuft zentral über OpenClaw Gateway (kein direct provider routing)
- API kommuniziert via `host.docker.internal:18789` (Gateway loopback — wird über `openclaw_default` Netzwerk gefixt)
- API kommuniziert über Docker-DNS via `http://openclaw-gateway-bao:18789` im gemeinsamen `openclaw_default`-Netzwerk.
- Der frühere Pfad `host.docker.internal:18789` ist auf dem VPS nicht erreichbar und wurde entfernt.
- **Achtung:** `[AllowAnonymous]` auf `/tasks/board` und `/tasks/reset-stale` muss durch ApiKey-Auth ersetzt werden (siehe [Architektur-Review](../docs/architecture-board-first-orchestration.md#61-kritisch--allowanonymous-auf-board-endpunkten))
- Vollständige Architektur- und Sicherheitsanalyse: [architecture-board-first-orchestration.md](../docs/architecture-board-first-orchestration.md)