unoverse 0.1.146 → 0.1.147

package/operator/ansible/ansible.cfg

@@ -17,3 +17,12 @@ become = True
 become_method = sudo
 become_user = root
 become_ask_pass = False
+
+[ssh_connection]
+# A transient TCP timeout must not kill a deploy: a single dropped SYN to a busy
+# or far-away VM read as UNREACHABLE on the very first task. Retry the connect
+# (ConnectionAttempts) and the connection (retries) before declaring the host gone.
+# The -C/ControlMaster/ControlPersist flags are ansible's defaults, restated
+# because ssh_args REPLACES them rather than appending.
+ssh_args = -C -o ControlMaster=auto -o ControlPersist=60s -o ConnectionAttempts=5 -o ConnectTimeout=15 -o ServerAliveInterval=15 -o ServerAliveCountMax=4
+retries = 3

package/operator/lib/deploy.sh

@@ -819,8 +819,20 @@ EOF
   # never been migrated, and reported success. The stamp is written only after install,
   # database and verify have all passed, so an interrupted setup resumes as a setup.
   if [ -z "$subcommand" ]; then
-    if ! ssh -o BatchMode=yes -o StrictHostKeyChecking=accept-new -o ConnectTimeout=15 \
-         "$deploy_user@$deploy_host" 'test -f /opt/gravity/.setup-complete' >/dev/null 2>&1; then
+    # "COULD NOT ASK" IS NOT "NOT SET UP". ssh exits 255 on a connection failure and
+    # with the remote command's status otherwise — and a transient timeout here once
+    # reclassified a healthy, deployed server as first-time, pointing a full install at
+    # it. Retry the connect, and if the host still cannot be reached, stop and say so
+    # rather than guessing what it is.
+    local stamp_probe=0
+    ssh -o BatchMode=yes -o StrictHostKeyChecking=accept-new -o ConnectTimeout=15 \
+        -o ConnectionAttempts=5 \
+        "$deploy_user@$deploy_host" 'test -f /opt/gravity/.setup-complete' >/dev/null 2>&1 || stamp_probe=$?
+    if [ "$stamp_probe" -eq 255 ]; then
+      rm -f "$tmp_inventory"
+      fail "Cannot reach $deploy_user@$deploy_host over SSH (tried 5 times). Check the network / security-group admin rule, then: unoverse deploy $cloud"
+      exit 1
+    elif [ "$stamp_probe" -ne 0 ]; then
       echo ""
       info "This server's setup has not finished. Running it ${DIM}(install, database, verify)${NC}"
       subcommand="first-time"

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "unoverse",
-  "version": "0.1.146",
+  "version": "0.1.147",
   "description": "The Unoverse front door — create a Studio project, a universe, or a client app, and launch Studio.",
   "license": "SEE LICENSE IN README.md",
   "type": "module",