redo

check_raid.yml

@@ -27,9 +27,12 @@
     raid_retries: "{{ RETRIES }}"
     raid_delay: 2
 
-    # SSH options (keeps same style as your working playbooks, but avoids auth prompts)
+    # Hard timeout for the whole SSH command (prevents hanging forever)
+    ssh_hard_timeout: 30
+
+    # SSH options (same style, but avoids auth prompts)
     ssh_opts:
-      - "-o"
+      - "-o"  # English comments
       - "StrictHostKeyChecking=no"
       - "-o"
       - "UserKnownHostsFile=/dev/null"
@@ -60,10 +63,6 @@
             print(f"ERROR: cannot read /proc/mdstat: {e}")
             sys.exit(2)
 
-        # Find token like [UU] / [U_] for the selected md device
-        # Example lines:
-        # md0 : active raid1 sdb1[0] sdc1[1]
-        #       11718751232 blocks super 1.2 [2/2] [UU]
         pat = re.compile(
             rf"^{re.escape(md)}\s*:\s*active.*\n\s*\d+\s+blocks.*\[\d+/\d+\]\s*\[([U_]+)\]",
             re.MULTILINE
@@ -76,8 +75,8 @@
 
         token = m.group(1)
         syncing = bool(re.search(r"\b(resync|recovery|reshape|check|repair)\b", txt))
-
         degraded = "_" in token
+
         print(f"RAID={md} token=[{token}] degraded={degraded} syncing={syncing} allow_sync={allow_sync}")
 
         if degraded:
@@ -98,30 +97,16 @@
 
     - name: Run RAID check commands on VM (via SSH)  # use SSHPASS env, hide item label
       ansible.builtin.command:
-        argv:
-          - sshpass
-          - -e
-          - ssh
-          - "{{ ssh_opts[0] }}"
-          - "{{ ssh_opts[1] }}"
-          - "{{ ssh_opts[2] }}"
-          - "{{ ssh_opts[3] }}"
-          - "{{ ssh_opts[4] }}"
-          - "{{ ssh_opts[5] }}"
-          - "{{ ssh_opts[6] }}"
-          - "{{ ssh_opts[7] }}"
-          - "{{ ssh_opts[8] }}"
-          - "{{ ssh_opts[9] }}"
-          - "{{ ssh_opts[10] }}"
-          - "{{ ssh_opts[11] }}"
-          - "{{ ssh_opts[12] }}"
-          - "{{ ssh_opts[13] }}"
-          - "{{ ssh_opts[14] }}"
-          - "{{ ssh_opts[15] }}"
-          - "{{ vm_user }}@{{ vm_ip }}"
-          - bash
-          - -lc
-          - "{{ ('sudo ' if use_sudo else '') + item }}"
+        argv: >-
+          {{
+            ['timeout', '-k', '5', (ssh_hard_timeout | string)]
+            + ['sshpass', '-e', 'ssh']
+            + ssh_opts
+            + [ vm_user ~ '@' ~ vm_ip,
+                'bash', '-lc',
+                ('sudo ' if use_sudo else '') + item
+              ]
+          }}
       environment:
         SSHPASS: "{{ vm_pass }}"
       loop: "{{ raid_commands }}"
@@ -134,7 +119,8 @@
       no_log: "{{ DEBUG == 0 }}"
       retries: "{{ raid_retries }}"
       delay: "{{ raid_delay }}"
-      until: raid_cmds is succeeded  # command executed (rc can be non-zero; we handle later)
+      # Retry only on typical SSH/timeout failures (255=ssh error, 124=timeout)
+      until: (raid_cmds.results | selectattr('rc', 'in', [124, 255]) | list | length) == 0
 
     - name: Show outputs for each RAID command
       ansible.builtin.debug: