DOCS-1594: tutorial for restoring database into RS

source/includes/toc-administration-backup-and-recovery.yaml

@@ -9,6 +9,11 @@ description: |
   system-level file snapshot tool, such as :term:`LVM` or native
   storage appliance tools.
 ---
+file: /tutorial/restore-database-from-backup-file
+description: |
+  Describes procedure for restoring a replica set from an archived
+  backup such as a :program:`mongodump` or `MMS Backup <mms.mongodb.com>`_ file.
+---
 file: /administration/backup-sharded-clusters
 description: |
   Detailed procedures and considerations for backing up sharded

source/tutorial/restore-database-from-backup-file.txt

@@ -0,0 +1,80 @@
+=====================================
+Restore a Database into a Replica Set
+=====================================
+
+.. default-domain:: mongodb
+
+Restore Database into a Single Node Replica Set
+-----------------------------------------------
+
+#. Obtain backup files, for example, from a :program:`mongodump`, or
+   an MMS Backup restore. See
+   :doc:`/tutorial/backup-databases-with-binary-database-dumps/` for
+   more about restoring with :program:`mongodump` or `Restore from a
+   Stored Snapshot
+   <http://mms.mongodb.com/help/backup/tutorial/restore-from-point-in-time-snapshot/>`_
+   in the MMS documentation for instructions detailing how to restore
+   using MMS Backup.
+
+#. Start a :program:`mongod` using your new data directory as its
+   ``dbpath``:
+
+   .. code-block:: sh
+
+      mongod --dbpath /data/db
+
+#. Convert your standalone :program:`mongod` process to a single node
+   replica set by shutting down the :program:`mongod` instance, and
+   restarting it with the ``--replSet`` option, as in the following
+   example:
+
+   .. code-block:: sh
+
+      mongod --dbpath /data/db --replSet <replName>
+
+#. Connect to the :program:`mongod` instance.
+
+#. User :method:`rs.initiate()` to initiate the new replica set.
+
+
+Add Members to the Replica Set
+------------------------------
+
+You restore the secondary members of a replica set in two ways: by
+manually copying the database files into their data directory, or
+through :ref:`initial sync <replica-set-initial-sync>`. The following
+sections provide instructions for each approach.
+
+.. note:: 
+
+   If your database is large, initial sync can take a long time to
+   complete. For large databases, it might be preferable to copy the
+   database files onto each host.
+
+Copy Database Files and Restart :program:`mongod` Instance
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+#. Shut down your :program:`mongod` instance.
+
+#. Copy your :term:`primary's <primary>` data directory (generally,
+   ``/data/db``) to the data directories of the other members of the
+   replica set. These will become :term:`secondaries <secondary>`.
+
+#. In a :program:`mongo` shell connected to the :term:`primary`, add
+   the :term:`secondaries <secondary>` to the replica set using
+   :method:`rs.add()`. For more information about deploying a replica
+   set, see: :doc:`/tutorial/deploy-replica-set`.
+
+Update Secondaries using Initial Sync
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+If you do not want to manually copy the data directories onto each
+member of your replica set, you can choose to add members to the
+replica set and catch them up using initial sync.
+
+#. Ensure that the data directories on the prospective replica set
+   members are empty.
+
+#. Add the members to the replica set. :ref:`Initial Sync
+   <replica-set-initial-sync>` will copy the data from the
+   :term:`primary` to the other members of the replica set.