Merge pull request 'fix: bug: disinto-edge crashes on cold disinto up — clones from forgejo before forgejo HTTP is ready (#665)' (#714) from fix/issue-665 into main

docker/edge/entrypoint-edge.sh

@@ -80,11 +80,29 @@ fi
 
 # Shallow clone at the pinned version — use clean URL, credential helper
 # supplies auth (#604).
+# Retry with exponential backoff — forgejo may still be starting (#665).
 if [ ! -d /opt/disinto/.git ]; then
   echo "edge: cloning ${FORGE_URL}/${FORGE_REPO} (branch ${DISINTO_VERSION:-main})..." >&2
-  if ! git clone --depth 1 --branch "${DISINTO_VERSION:-main}" "${FORGE_URL}/${FORGE_REPO}.git" /opt/disinto; then
+  _clone_ok=false
+  _backoff=2
+  _max_backoff=30
+  _max_attempts=10
+  for _attempt in $(seq 1 "$_max_attempts"); do
+    if git clone --depth 1 --branch "${DISINTO_VERSION:-main}" "${FORGE_URL}/${FORGE_REPO}.git" /opt/disinto 2>&1; then
+      _clone_ok=true
+      break
+    fi
+    rm -rf /opt/disinto  # clean up partial clone before retry
+    if [ "$_attempt" -lt "$_max_attempts" ]; then
+      echo "edge: clone attempt ${_attempt}/${_max_attempts} failed, retrying in ${_backoff}s..." >&2
+      sleep "$_backoff"
+      _backoff=$(( _backoff * 2 ))
+      if [ "$_backoff" -gt "$_max_backoff" ]; then _backoff=$_max_backoff; fi
+    fi
+  done
+  if [ "$_clone_ok" != "true" ]; then
     echo >&2
-    echo "FATAL: failed to clone ${FORGE_URL}/${FORGE_REPO}.git (branch ${DISINTO_VERSION:-main})" >&2
+    echo "FATAL: failed to clone ${FORGE_URL}/${FORGE_REPO}.git (branch ${DISINTO_VERSION:-main}) after ${_max_attempts} attempts" >&2
     echo "Likely causes:" >&2
     echo "  - Forgejo at ${FORGE_URL} is unreachable from the edge container" >&2
     echo "  - Repository '${FORGE_REPO}' does not exist on this forge" >&2

lib/generators.sh

@@ -143,8 +143,10 @@ _generate_local_model_services() {
       ARCHITECT_INTERVAL: "${ARCHITECT_INTERVAL:-21600}"
       PLANNER_INTERVAL: "${PLANNER_INTERVAL:-43200}"
     depends_on:
-      - forgejo
-      - woodpecker
+      forgejo:
+        condition: service_healthy
+      woodpecker:
+        condition: service_started
     networks:
       - disinto-net
     profiles: ["agents-${service_name}"]
@@ -263,6 +265,12 @@ services:
       FORGEJO__security__INSTALL_LOCK: "true"
       FORGEJO__service__DISABLE_REGISTRATION: "true"
       FORGEJO__webhook__ALLOWED_HOST_LIST: "private"
+    healthcheck:
+      test: ["CMD", "wget", "-q", "--spider", "http://localhost:3000/api/v1/version"]
+      interval: 5s
+      timeout: 3s
+      retries: 30
+      start_period: 30s
     networks:
       - disinto-net
 
@@ -289,7 +297,8 @@ services:
       WOODPECKER_DATABASE_DATASOURCE: /var/lib/woodpecker/woodpecker.sqlite
       WOODPECKER_ENVIRONMENT: "FORGE_TOKEN:${FORGE_TOKEN}"
     depends_on:
-      - forgejo
+      forgejo:
+        condition: service_healthy
     networks:
       - disinto-net
 
@@ -364,8 +373,10 @@ services:
     # .env.vault.enc and are NEVER injected here — only the runner
     # container receives them at fire time (AD-006, #745).
     depends_on:
-      - forgejo
-      - woodpecker
+      forgejo:
+        condition: service_healthy
+      woodpecker:
+        condition: service_started
     networks:
       - disinto-net
 
@@ -426,9 +437,12 @@ services:
       - ${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}:${CLAUDE_SHARED_DIR:-/var/lib/disinto/claude-shared}
       - ${HOME}/.claude.json:/home/agent/.claude.json:ro
     depends_on:
-      - forgejo
-      - woodpecker
-      - staging
+      forgejo:
+        condition: service_healthy
+      woodpecker:
+        condition: service_started
+      staging:
+        condition: service_started
     networks:
       - disinto-net