SERVER-62065 Introduce the 'repairShardedCollectionChunksHistory' command

(cherry picked from commit 146c18b9954abc116046621dc4849cc7d97ef523)

6 files changed, 128 insertions, 4 deletions

diff --git a/src/mongo/s/catalog/type_chunk.cpp b/src/mongo/s/catalog/type_chunk.cpp
index 30e814256c9..ad1430685a8 100644
--- a/src/mongo/s/catalog/type_chunk.cpp
+++ b/src/mongo/s/catalog/type_chunk.cpp
@@ -59,6 +59,7 @@ const BSONField<Date_t> ChunkType::lastmod("lastmod");
 const BSONField<OID> ChunkType::epoch("lastmodEpoch");
 const BSONField<Timestamp> ChunkType::timestamp("lastmodTimestamp");
 const BSONField<BSONObj> ChunkType::history("history");
+const BSONField<bool> ChunkType::historyIsAt40("historyIsAt40");
 
 namespace {
 
diff --git a/src/mongo/s/catalog/type_chunk.h b/src/mongo/s/catalog/type_chunk.h
index 778d3af77aa..4f3d9177074 100644
--- a/src/mongo/s/catalog/type_chunk.h
+++ b/src/mongo/s/catalog/type_chunk.h
@@ -212,6 +212,7 @@ public:
     static const BSONField<OID> epoch;
     static const BSONField<Timestamp> timestamp;
     static const BSONField<BSONObj> history;
+    static const BSONField<bool> historyIsAt40;
 
     ChunkType();
     ChunkType(NamespaceString nss, ChunkRange range, ChunkVersion version, ShardId shardId);
diff --git a/src/mongo/s/client/shard.h b/src/mongo/s/client/shard.h
index 9c28d5139c8..285aaf48994 100644
--- a/src/mongo/s/client/shard.h
+++ b/src/mongo/s/client/shard.h
@@ -41,11 +41,11 @@
 #include "mongo/db/repl/read_concern_args.h"
 #include "mongo/executor/remote_command_response.h"
 #include "mongo/s/shard_id.h"
+#include "mongo/s/write_ops/batched_command_request.h"
+#include "mongo/s/write_ops/batched_command_response.h"
 
 namespace mongo {
 
-class BatchedCommandRequest;
-class BatchedCommandResponse;
 class OperationContext;
 class RemoteCommandTargeter;
 
diff --git a/src/mongo/s/commands/SConscript b/src/mongo/s/commands/SConscript
index dd39115b35b..0fb51bc21c6 100644
--- a/src/mongo/s/commands/SConscript
+++ b/src/mongo/s/commands/SConscript
@@ -84,6 +84,7 @@ env.Library(
         'cluster_remove_shard_cmd.cpp',
         'cluster_remove_shard_from_zone_cmd.cpp',
         'cluster_rename_collection_cmd.cpp',
+        'cluster_repair_sharded_collection_chunks_history_cmd.cpp',
         'cluster_repl_set_get_status_cmd.cpp',
         'cluster_reshard_collection_cmd.cpp',
         'cluster_rwc_defaults_commands.cpp',
diff --git a/src/mongo/s/commands/cluster_repair_sharded_collection_chunks_history_cmd.cpp b/src/mongo/s/commands/cluster_repair_sharded_collection_chunks_history_cmd.cpp
new file mode 100644
index 00000000000..dd08b1574d4
--- /dev/null
+++ b/src/mongo/s/commands/cluster_repair_sharded_collection_chunks_history_cmd.cpp
@@ -0,0 +1,121 @@
+/**
+ *    Copyright (C) 2021-present MongoDB, Inc.
+ *
+ *    This program is free software: you can redistribute it and/or modify
+ *    it under the terms of the Server Side Public License, version 1,
+ *    as published by MongoDB, Inc.
+ *
+ *    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
+ *    Server Side Public License for more details.
+ *
+ *    You should have received a copy of the Server Side Public License
+ *    along with this program. If not, see
+ *    <http://www.mongodb.com/licensing/server-side-public-license>.
+ *
+ *    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 Server Side 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.
+ */
+
+#define MONGO_LOG_DEFAULT_COMPONENT ::mongo::logger::LogComponent::kSharding
+
+#include "mongo/platform/basic.h"
+
+#include "mongo/db/audit.h"
+#include "mongo/db/auth/action_set.h"
+#include "mongo/db/auth/action_type.h"
+#include "mongo/db/auth/authorization_manager.h"
+#include "mongo/db/auth/authorization_session.h"
+#include "mongo/db/client.h"
+#include "mongo/db/commands.h"
+#include "mongo/s/grid.h"
+
+namespace mongo {
+namespace {
+
+class RepairShardedCollectionChunksHistoryCommand : public BasicCommand {
+public:
+    RepairShardedCollectionChunksHistoryCommand()
+        : BasicCommand("repairShardedCollectionChunksHistory") {}
+
+    AllowedOnSecondary secondaryAllowed(ServiceContext*) const override {
+        return AllowedOnSecondary::kAlways;
+    }
+
+    bool adminOnly() const override {
+        return true;
+    }
+
+    bool supportsWriteConcern(const BSONObj& cmd) const override {
+        return false;
+    }
+
+    std::string help() const override {
+        return "Administrative command to repair the effects of SERVER-62065. If the collection "
+               "has been upgraded through a cluster comprised of binaries which do not contain "
+               "this command, the chunks cache collections on the shards will miss history "
+               "entries. This command will correct that and will mark such collections as "
+               "correctly repaired, so that a subsequent invocation will not cause any changes to "
+               "the routing information. In rare cases where the history entries are missing due "
+               "to corrupted restore, the 'force:true' parameter can be passed which will force "
+               "all history entries to be re-added.";
+    }
+
+    // The command intentionally uses the permission control of split/mergeChunks since it only
+    // modifies the contents of chunk entries and increments the collection/shard versions without
+    // causing any data placement changes
+    Status checkAuthForCommand(Client* client,
+                               const std::string& dbname,
+                               const BSONObj& cmdObj) const override {
+        if (!AuthorizationSession::get(client)->isAuthorizedForActionsOnResource(
+                ResourcePattern::forExactNamespace(NamespaceString(parseNs(dbname, cmdObj))),
+                ActionType::splitChunk)) {
+            return Status(ErrorCodes::Unauthorized, "Unauthorized");
+        }
+        return Status::OK();
+    }
+
+    std::string parseNs(const std::string& unusedDbName, const BSONObj& cmdObj) const override {
+        return CommandHelpers::parseNsFullyQualified(cmdObj);
+    }
+
+    bool run(OperationContext* opCtx,
+             const std::string& unusedDbName,
+             const BSONObj& cmdObj,
+             BSONObjBuilder& result) override {
+        const NamespaceString nss{parseNs(unusedDbName, cmdObj)};
+
+        BSONObjBuilder cmdBuilder(
+            BSON("_configsvrRepairShardedCollectionChunksHistory" << nss.ns()));
+        if (cmdObj["force"].booleanSafe())
+            cmdBuilder.appendBool("force", true);
+
+        auto configShard = Grid::get(opCtx)->shardRegistry()->getConfigShard();
+        auto cmdResponse = uassertStatusOK(configShard->runCommandWithFixedRetryAttempts(
+            opCtx,
+            ReadPreferenceSetting{ReadPreference::PrimaryOnly},
+            "admin",
+            CommandHelpers::appendMajorityWriteConcern(cmdBuilder.obj(), opCtx->getWriteConcern()),
+            Shard::RetryPolicy::kIdempotent));
+        uassertStatusOK(cmdResponse.commandStatus);
+
+        // Append any return value from the response, which the config server returned
+        CommandHelpers::filterCommandReplyForPassthrough(cmdResponse.response, &result);
+
+        return true;
+    }
+
+} repairShardedCollectionChunksHistoryCommand;
+
+}  // namespace
+}  // namespace mongo
diff --git a/src/mongo/s/request_types/flush_routing_table_cache_updates.idl b/src/mongo/s/request_types/flush_routing_table_cache_updates.idl
index 3548b429c3a..62a8e3a94a5 100644
--- a/src/mongo/s/request_types/flush_routing_table_cache_updates.idl
+++ b/src/mongo/s/request_types/flush_routing_table_cache_updates.idl
@@ -26,8 +26,6 @@
 # it in the license file.
 #
 
-# _flushRoutingTableCacheUpdates IDL File
-
 global:
     cpp_namespace: "mongo"
 
@@ -40,6 +38,7 @@ commands:
             An internal command to wait for the last routing table cache refresh for a
             particular namespace to be persisted to disk.
         command_name: _flushRoutingTableCacheUpdates
+        cpp_name: FlushRoutingTableCacheUpdates
         strict: true
         namespace: type
         api_version: ""
@@ -55,6 +54,7 @@ commands:
         description: >-
             The same behavior as _flushRoutingTableCacheUpdates but accepts writeConcern.
         command_name: _flushRoutingTableCacheUpdatesWithWriteConcern
+        cpp_name: FlushRoutingTableCacheUpdatesWithWriteConcern
         strict: true
         namespace: type
         api_version: ""