Allow "bin/repository thaw --demote" to demote an entire service, not just a single device

Summary: Ref T13222. See PHI992. If you lose an entire cluster, you may want to aggressively demote it out of existence. You currently need to `xargs` your way through this. Allow `--demote <service>`, which demotes all devices in a service.

Test Plan: Demoted with `--demote <device>` and `--demote <service>`. Hit the `--promote service` error.

Reviewers: amckinley

Reviewed By: amckinley

Maniphest Tasks: T13222

Differential Revision: https://secure.phabricator.com/D19850

src/applications/repository/management/PhabricatorRepositoryManagementThawWorkflow.php

@@ -15,10 +15,11 @@ final class PhabricatorRepositoryManagementThawWorkflow
         array(
           array(
             'name' => 'demote',
-            'param' => 'device',
+            'param' => 'device/service',
             'help' => pht(
-              'Demote a device, discarding local changes. Clears stuck '.
+              'Demote a device (or all devices in a service) discarding '.
-              'write locks and recovers from lost leaders.'),
+              'local changes. Clears stuck write locks and recovers from '.
+              'lost leaders.'),
           ),
           array(
             'name' => 'promote',
@@ -61,15 +62,53 @@ final class PhabricatorRepositoryManagementThawWorkflow
         pht('Specify either --promote or --demote, but not both.'));
     }
 
-    $device_name = nonempty($promote, $demote);
+    $target_name = nonempty($promote, $demote);
 
-    $device = id(new AlmanacDeviceQuery())
+    $devices = id(new AlmanacDeviceQuery())
       ->setViewer($viewer)
-      ->withNames(array($device_name))
+      ->withNames(array($target_name))
+      ->execute();
+    if (!$devices) {
+      $service = id(new AlmanacServiceQuery())
+        ->setViewer($viewer)
+        ->withNames(array($target_name))
         ->executeOne();
-    if (!$device) {
+
+      if (!$service) {
         throw new PhutilArgumentUsageException(
-        pht('No device "%s" exists.', $device_name));
+          pht('No device or service named "%s" exists.', $target_name));
+      }
+
+      if ($promote) {
+        throw new PhutilArgumentUsageException(
+          pht(
+            'You can not "--promote" an entire service ("%s"). Only a single '.
+            'device may be promoted.',
+            $target_name));
+      }
+
+      $bindings = id(new AlmanacBindingQuery())
+        ->setViewer($viewer)
+        ->withServicePHIDs(array($service->getPHID()))
+        ->execute();
+      if (!$bindings) {
+        throw new PhutilArgumentUsageException(
+          pht(
+            'Service "%s" is not bound to any devices.',
+            $target_name));
+      }
+
+      $interfaces = id(new AlmanacInterfaceQuery())
+        ->setViewer($viewer)
+        ->withPHIDs(mpull($bindings, 'getInterfacePHID'))
+        ->execute();
+
+      $device_phids = mpull($interfaces, 'getDevicePHID');
+
+      $devices = id(new AlmanacDeviceQuery())
+        ->setViewer($viewer)
+        ->withPHIDs($device_phids)
+        ->execute();
     }
 
     $repository_names = $args->getArg('repositories');
@@ -97,7 +136,7 @@ final class PhabricatorRepositoryManagementThawWorkflow
 
       $services = id(new AlmanacServiceQuery())
         ->setViewer($viewer)
-        ->withDevicePHIDs(array($device->getPHID()))
+        ->withDevicePHIDs(mpull($devices, 'getPHID'))
         ->execute();
       if ($services) {
         $repositories = id(new PhabricatorRepositoryQuery())
@@ -108,7 +147,7 @@ final class PhabricatorRepositoryManagementThawWorkflow
 
       if (!$repositories) {
         throw new PhutilArgumentUsageException(
-          pht('There are no repositories on the selected device.'));
+          pht('There are no repositories on the selected device or service.'));
       }
     }
 
@@ -150,6 +189,7 @@ final class PhabricatorRepositoryManagementThawWorkflow
         pht('User aborted the workflow.'));
     }
 
+    foreach ($devices as $device) {
       foreach ($repositories as $repository) {
         $repository_phid = $repository->getPHID();
 
@@ -182,8 +222,8 @@ final class PhabricatorRepositoryManagementThawWorkflow
             if (empty($bindings[$device->getPHID()])) {
               throw new PhutilArgumentUsageException(
                 pht(
-                'Repository "%s" has no active binding to device "%s". Only '.
+                  'Repository "%s" has no active binding to device "%s". '.
-                'actively bound devices can be promoted.',
+                  'Only actively bound devices can be promoted.',
                   $repository->getDisplayName(),
                   $device->getName()));
             }
@@ -192,9 +232,9 @@ final class PhabricatorRepositoryManagementThawWorkflow
               $repository->getPHID());
             $versions = mpull($versions, null, 'getDevicePHID');
 
-          // Before we promote, make sure there are no outstanding versions on
+            // Before we promote, make sure there are no outstanding versions
-          // devices with inactive bindings. If there are, you need to demote
+            // on devices with inactive bindings. If there are, you need to
-          // these first.
+            // demote these first.
             $inactive = array();
             foreach ($versions as $device_phid => $version) {
               if (isset($bindings[$device_phid])) {
@@ -220,8 +260,8 @@ final class PhabricatorRepositoryManagementThawWorkflow
             }
 
             // Now, make sure there are no outstanding versions on devices with
-          // active bindings. These also need to be demoted (or promoting is a
+            // active bindings. These also need to be demoted (or promoting is
-          // mistake or already happened).
+            // a mistake or already happened).
             $active = array_select_keys($versions, array_keys($bindings));
             if ($active) {
               $handles = $viewer->loadHandles(array_keys($active));
@@ -271,6 +311,7 @@ final class PhabricatorRepositoryManagementThawWorkflow
 
         $write_lock->unlock();
       }
+    }
 
     return 0;
   }

src/docs/user/cluster/cluster_repositories.diviner

@@ -433,6 +433,18 @@ If you do this, **you will lose unreplicated data**. You will discard any
 changes on the affected leaders which have not replicated to other devices
 in the cluster.
 
+If you have lost an entire cluster and replaced it with new devices that you
+have restored from backups, you can aggressively wipe all memory of the old
+devices by using `--demote <service>` and `--all-repositories`. **This is
+dangerous and discards all unreplicated data in any repository on any device.**
+
+```
+phabricator/ $ ./bin/repository thaw --demote repo.corp.net --all-repositories
+```
+
+After you do this, continue below to promote a leader and restore the cluster
+to service.
+
 
 Ambiguous Leaders
 =================