SERVER-33206 Create MovePrimarySourceManager and add clone command

1 files changed, 146 insertions, 0 deletions

diff --git a/src/mongo/db/s/move_primary_source_manager.h b/src/mongo/db/s/move_primary_source_manager.h
new file mode 100644
index 00000000000..4c9eb5a63c9
--- /dev/null
+++ b/src/mongo/db/s/move_primary_source_manager.h
@@ -0,0 +1,146 @@
+/**
+ *    Copyright (C) 2018 MongoDB Inc.
+ *
+ *    This program is free software: you can redistribute it and/or  modify
+ *    it under the terms of the GNU Affero General Public License, version 3,
+ *    as published by the Free Software Foundation.
+ *
+ *    This program is distributed in the hope that it will be useful,
+ *    but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *    GNU Affero General Public License for more details.
+ *
+ *    You should have received a copy of the GNU Affero General Public License
+ *    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ *    As a special exception, the copyright holders give permission to link the
+ *    code of portions of this program with the OpenSSL library under certain
+ *    conditions as described in each individual source file and distribute
+ *    linked combinations including the program with the OpenSSL library. You
+ *    must comply with the GNU Affero General Public License in all respects
+ *    for all of the code used other than as permitted herein. If you modify
+ *    file(s) with this exception, you may extend this exception to your
+ *    version of the file(s), but you are not obligated to do so. If you do not
+ *    wish to do so, delete this exception statement from your version. If you
+ *    delete this exception statement from all source files in the program,
+ *    then also delete it in the license file.
+ */
+
+#pragma once
+
+#include "mongo/base/disallow_copying.h"
+#include "mongo/db/s/database_sharding_state.h"
+#include "mongo/s/request_types/move_primary_gen.h"
+#include "mongo/util/timer.h"
+
+namespace mongo {
+
+class OperationContext;
+class Shard;
+struct ShardingStatistics;
+class Status;
+
+/**
+ * The donor-side movePrimary state machine. This object must be created and owned by a single
+ * thread, which controls its lifetime and should not be passed across threads. Unless explicitly
+ * indicated its methods must not be called from more than one thread and must not be called while
+ * any locks are held.
+ *
+ * The intended workflow is as follows:
+ *  - Acquire a distributed lock on the database whose primary is about to be moved.
+ *  - Instantiate a MovePrimarySourceManager on the stack.
+ *  - Call clone to start and finish cloning of the unsharded collections.
+ *  - Call updatePrimary to indicate the new primary in the config server metadata.
+ *  - Call clearStaleCollections to drop now-unused collections (and potentially databases) on the
+ *      old primary.
+ *
+ * At any point in time it is safe to let the MovePrimarySourceManager object go out of scope in
+ * which case the destructor will take care of clean up based on how far we have advanced.
+ */
+class MovePrimarySourceManager {
+    MONGO_DISALLOW_COPYING(MovePrimarySourceManager);
+
+public:
+    /**
+     * Instantiates a new movePrimary source manager. Must be called with the distributed lock
+     * acquired in advance (not asserted).
+     *
+     * May throw any exception. Known exceptions (TODO) are:
+     *  - InvalidOptions if the operation context is missing database version
+     *  - StaleConfigException if the expected database version does not match what we find it
+     *      to be after acquiring the distributed lock.
+     */
+
+    MovePrimarySourceManager(OperationContext* opCtx,
+                             ShardMovePrimary requestArgs,
+                             StringData dbname,
+                             std::unique_ptr<Shard> fromShard,
+                             std::unique_ptr<Shard> toShard);
+    ~MovePrimarySourceManager();
+
+    /**
+     * Returns the namespace for which this source manager is active.
+     */
+    NamespaceString getNss() const;
+
+    /**
+     * Contacts the recipient shard and tells it to start cloning the specified chunk. This method
+     * will fail if for any reason the recipient shard fails to complete the cloning sequence.
+     *
+     * Expected state: kCreated
+     * Resulting state: kCloning on success, kDone on failure
+     */
+    Status clone(OperationContext* opCtx);
+
+    /**
+     * May be called at any time. Unregisters the movePrimary source manager from the database and
+     * logs an error in the change log to indicate that the migration has failed.
+     *
+     * Expected state: Any
+     * Resulting state: kDone
+     */
+    void cleanupOnError(OperationContext* opCtx);
+
+private:
+    static BSONObj _buildMoveLogEntry(const std::string& db,
+                                      const std::string& from,
+                                      const std::string& to) {
+        BSONObjBuilder details;
+        details.append("database", db);
+        details.append("from", from);
+        details.append("to", to);
+
+        return details.obj();
+    }
+
+    // Used to track the current state of the source manager. See the methods above, which have
+    // comments explaining the various state transitions.
+    // TODO: Will tweak these as further work on moving movePrimary to shards
+    enum State { kCreated, kCloning, kCloneCompleted, kCriticalSection, kDone };
+
+    /**
+     * Called when any of the states fails. May only be called once and will put the migration
+     * manager into the kDone state.
+     */
+    void _cleanup(OperationContext* opCtx);
+
+    // The parameters to the movePrimary command
+    const ShardMovePrimary _requestArgs;
+
+    // The database whose primary we are moving.
+    const StringData _dbname;
+
+    // The donor shard
+    const std::unique_ptr<Shard> _fromShard;
+
+    // The recipient shard
+    const std::unique_ptr<Shard> _toShard;
+
+    // Times the entire movePrimary operation
+    const Timer _entireOpTimer;
+
+    // The current state. Used only for diagnostics and validation.
+    State _state{kCreated};
+};
+
+}  // namespace mongo