repmgr: have "cluster show" exit with a non-zero value if issues detected

If any issues are detected (e.g. node not reachable, unexpected node status
etc.), "repmgr cluster show" returns exit code 25 ("ERR_NODE_STATUS").

Note that exit code 25 was introduced recently as "ERR_CLUSTER_CHECK",
however it makes sense to use this to indicate issues detected by any
command which can detect node issues.

Addresses GitHub #456.

HISTORY

@@ -2,6 +2,8 @@
 		repmgr: add "--missing-slots" check to "repmgr node check" (Ian)
         repmgr: improve command line error handling; GitHub #464 (Ian)
         repmgr: fix "standby register --wait-sync" when no timeout provided (Ian)
+        repmgr: "cluster show" returns non-zero value if an issue encountered;
+          GitHub #456 (Ian)
         repmgrd: create a PID file by default; GitHub #457 (Ian)
 		repmgrd: daemonize process by default; GitHub #458 (Ian)
 

doc/appendix-release-notes.sgml

@@ -46,6 +46,12 @@
             </para>
           </listitem>
 
+          <listitem>
+            <para>
+              <command><link linkend="repmgr-cluster-show">repmgr cluster-show</link></command>
+              returns non-zero exit code if node status issues detected (GitHub #456).
+            </para>
+          </listitem>
 
           <listitem>
             <para>
@@ -76,7 +82,7 @@
 
           <listitem>
             <para>
-              <application>repmgr</application>: fix <command><link linkend="repmgr-standby-register">repmgr standby register--wait-sync </link></command>
+              <application>repmgr</application>: fix <command><link linkend="repmgr-standby-register">repmgr standby register --wait-sync</link></command>
               when no timeout provided.
             </para>
           </listitem>

doc/repmgr-cluster-crosscheck.sgml

@@ -56,7 +56,7 @@
       </varlistentry>
 
       <varlistentry>
-        <term><option>ERR_CLUSTER_CHECK (25)</option></term>
+        <term><option>ERR_NODE_STATUS (25)</option></term>
         <listitem>
           <para>
             One or more nodes could not be reached.

doc/repmgr-cluster-matrix.sgml

@@ -116,7 +116,7 @@
       </varlistentry>
 
       <varlistentry>
-        <term><option>ERR_CLUSTER_CHECK (25)</option></term>
+        <term><option>ERR_NODE_STATUS (25)</option></term>
         <listitem>
           <para>
             One or more nodes could not be reached.

doc/repmgr-cluster-show.sgml

@@ -113,4 +113,33 @@
     </para>
   </refsect1>
 
+
+  <refsect1>
+    <title>Exit codes</title>
+    <para>
+      Following exit codes can be emitted by <command>repmgr cluster show</command>:
+    </para>
+    <variablelist>
+
+      <varlistentry>
+        <term><option>SUCCESS (0)</option></term>
+        <listitem>
+          <para>
+            No issues were detected.
+          </para>
+        </listitem>
+      </varlistentry>
+
+      <varlistentry>
+        <term><option>ERR_NODE_STATUS (25)</option></term>
+        <listitem>
+          <para>
+            One or more issues were detected.
+          </para>
+        </listitem>
+      </varlistentry>
+
+   </variablelist>
+  </refsect1>
+
 </refentry>

errcode.h

@@ -46,6 +46,6 @@
 #define ERR_SWITCHOVER_INCOMPLETE 22
 #define ERR_FOLLOW_FAIL 23
 #define ERR_REJOIN_FAIL 24
-#define ERR_CLUSTER_CHECK 25
+#define ERR_NODE_STATUS 25
 
 #endif							/* _ERRCODE_H_ */

repmgr-action-cluster.c

@@ -83,6 +83,7 @@ do_cluster_show(void)
 	int			i = 0;
 	ItemList	warnings = {NULL, NULL};
 	bool		success = false;
+	bool		error_found = false;
 
 	/* Connect to local database to obtain cluster connection data */
 	log_verbose(LOG_INFO, _("connecting to database"));
@@ -218,6 +219,7 @@ do_cluster_show(void)
 						else
 						{
 							appendPQExpBuffer(&details, "- failed");
+							error_found = true;
 						}
 					}
 				}
@@ -281,6 +283,7 @@ do_cluster_show(void)
 						else
 						{
 							appendPQExpBuffer(&details, "- failed");
+							error_found = true;
 						}
 					}
 				}
@@ -292,17 +295,27 @@ do_cluster_show(void)
 					if (cell->node_info->node_status == NODE_STATUS_UP)
 					{
 						if (cell->node_info->active == true)
+						{
 							appendPQExpBuffer(&details, "* running");
+						}
 						else
+						{
 							appendPQExpBuffer(&details, "! running");
+							error_found = true;
+						}
 					}
 					/* node is unreachable */
 					else
 					{
 						if (cell->node_info->active == true)
+						{
 							appendPQExpBuffer(&details, "? unreachable");
+						}
 						else
+						{
 							appendPQExpBuffer(&details, "- failed");
+							error_found = true;
+						}
 					}
 				}
 				break;
@@ -310,6 +323,7 @@ do_cluster_show(void)
 				{
 					/* this should never happen */
 					appendPQExpBuffer(&details, "? unknown node type");
+						error_found = true;
 				}
 				break;
 		}
@@ -414,7 +428,6 @@ do_cluster_show(void)
 	PQfinish(conn);
 
 	/* emit any warnings */
-
 	if (warnings.head != NULL && runtime_options.terse == false && runtime_options.output_mode != OM_CSV)
 	{
 		ItemListCell *cell = NULL;
@@ -425,6 +438,20 @@ do_cluster_show(void)
 			printf(_("  - %s\n"), cell->string);
 		}
 	}
+
+	/*
+	 * If warnings were noted, even if they're not displayed (e.g. in --csv node),
+	 * that means something's not right so we need to emit a non-zero exit code.
+	 */
+	if (warnings.head != NULL)
+	{
+		error_found = true;
+	}
+
+	if (error_found == true)
+	{
+		exit(ERR_NODE_STATUS);
+	}
 }
 
 
@@ -696,7 +723,7 @@ do_cluster_crosscheck(void)
 
 	if (error_found == true)
 	{
-		exit(ERR_CLUSTER_CHECK);
+		exit(ERR_NODE_STATUS);
 	}
 }
 
@@ -786,7 +813,7 @@ do_cluster_matrix()
 
 	if (error_found == true)
 	{
-		exit(ERR_CLUSTER_CHECK);
+		exit(ERR_NODE_STATUS);
 	}
 }