summaryrefslogtreecommitdiff
path: root/src/mongo/s/client/shard_local.cpp
blob: 5270842a8bad084b30fed8ff2688220c5189f223 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
/**
 * Copyright (C) 2016 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.
 */

#define MONGO_LOG_DEFAULT_COMPONENT ::mongo::logger::LogComponent::kSharding

#include <boost/none_t.hpp>

#include "mongo/platform/basic.h"

#include "mongo/s/client/shard_local.h"

#include "mongo/base/status.h"
#include "mongo/base/status_with.h"
#include "mongo/client/remote_command_targeter.h"
#include "mongo/db/curop.h"
#include "mongo/db/dbdirectclient.h"
#include "mongo/db/repl/repl_client_info.h"
#include "mongo/db/repl/repl_set_config.h"
#include "mongo/db/repl/replication_coordinator_global.h"
#include "mongo/rpc/get_status_from_command_result.h"
#include "mongo/rpc/unique_message.h"
#include "mongo/util/log.h"
#include "mongo/util/scopeguard.h"

namespace mongo {

ShardLocal::ShardLocal(const ShardId& id) : Shard(id) {
    // Currently ShardLocal only works for config servers. If we ever start using ShardLocal on
    // shards we'll need to consider how to handle shards.
    invariant(serverGlobalParams.clusterRole == ClusterRole::ConfigServer);
}

const ConnectionString ShardLocal::getConnString() const {
    MONGO_UNREACHABLE;
}

std::shared_ptr<RemoteCommandTargeter> ShardLocal::getTargeter() const {
    MONGO_UNREACHABLE;
};

const ConnectionString ShardLocal::originalConnString() const {
    // Return the local connection string here as this method is only used for updating the
    // ShardRegistry and we don't need a mapping from hosts in the replica set config to the shard
    // for local shards.
    return ConnectionString::forLocal();
}

void ShardLocal::updateReplSetMonitor(const HostAndPort& remoteHost,
                                      const Status& remoteCommandStatus) {
    MONGO_UNREACHABLE;
}

std::string ShardLocal::toString() const {
    return getId().toString() + ":<local>";
}

bool ShardLocal::isRetriableError(ErrorCodes::Error code, RetryPolicy options) {
    if (options == RetryPolicy::kNoRetry) {
        return false;
    }

    if (options == RetryPolicy::kIdempotent) {
        return code == ErrorCodes::WriteConcernFailed;
    } else {
        invariant(options == RetryPolicy::kNotIdempotent);
        return false;
    }
}

void ShardLocal::_updateLastOpTimeFromClient(OperationContext* opCtx,
                                             const repl::OpTime& previousOpTimeOnClient) {
    repl::OpTime lastOpTimeFromClient =
        repl::ReplClientInfo::forClient(opCtx->getClient()).getLastOp();
    invariant(lastOpTimeFromClient >= previousOpTimeOnClient);
    if (lastOpTimeFromClient.isNull() || lastOpTimeFromClient == previousOpTimeOnClient) {
        return;
    }

    stdx::lock_guard<stdx::mutex> lk(_mutex);
    if (lastOpTimeFromClient >= _lastOpTime) {
        // It's always possible for lastOpTimeFromClient to be less than _lastOpTime if another
        // thread started and completed a write through this ShardLocal (updating _lastOpTime)
        // after this operation had completed its write but before it got here.
        _lastOpTime = lastOpTimeFromClient;
    }
}

repl::OpTime ShardLocal::_getLastOpTime() {
    stdx::lock_guard<stdx::mutex> lk(_mutex);
    return _lastOpTime;
}

Shard::HostWithResponse ShardLocal::_runCommand(OperationContext* opCtx,
                                                const ReadPreferenceSetting& unused,
                                                const std::string& dbName,
                                                Milliseconds maxTimeMSOverrideUnused,
                                                const BSONObj& cmdObj) {
    repl::OpTime currentOpTimeFromClient =
        repl::ReplClientInfo::forClient(opCtx->getClient()).getLastOp();
    ON_BLOCK_EXIT([this, &opCtx, &currentOpTimeFromClient] {
        _updateLastOpTimeFromClient(opCtx, currentOpTimeFromClient);
    });

    try {
        DBDirectClient client(opCtx);
        rpc::UniqueReply commandResponse = client.runCommandWithMetadata(
            dbName, cmdObj.firstElementFieldName(), rpc::makeEmptyMetadata(), cmdObj);
        BSONObj responseReply = commandResponse->getCommandReply().getOwned();
        BSONObj responseMetadata = commandResponse->getMetadata().getOwned();

        Status commandStatus = getStatusFromCommandResult(responseReply);
        Status writeConcernStatus = getWriteConcernStatusFromCommandResult(responseReply);

        return Shard::HostWithResponse(boost::none,
                                       Shard::CommandResponse{std::move(responseReply),
                                                              std::move(responseMetadata),
                                                              std::move(commandStatus),
                                                              std::move(writeConcernStatus)});
    } catch (const DBException& ex) {
        return Shard::HostWithResponse(boost::none, ex.toStatus());
    }
}

StatusWith<Shard::QueryResponse> ShardLocal::_exhaustiveFindOnConfig(
    OperationContext* opCtx,
    const ReadPreferenceSetting& readPref,
    const repl::ReadConcernLevel& readConcernLevel,
    const NamespaceString& nss,
    const BSONObj& query,
    const BSONObj& sort,
    boost::optional<long long> limit) {
    auto replCoord = repl::ReplicationCoordinator::get(opCtx);

    if (readConcernLevel == repl::ReadConcernLevel::kMajorityReadConcern) {
        // Set up operation context with majority read snapshot so correct optime can be retrieved.
        Status status = opCtx->recoveryUnit()->setReadFromMajorityCommittedSnapshot();

        // Wait for any writes performed by this ShardLocal instance to be committed and visible.
        Status readConcernStatus = replCoord->waitUntilOpTimeForRead(
            opCtx, repl::ReadConcernArgs{_getLastOpTime(), readConcernLevel});
        if (!readConcernStatus.isOK()) {
            return readConcernStatus;
        }

        // Inform the storage engine to read from the committed snapshot for the rest of this
        // operation.
        status = opCtx->recoveryUnit()->setReadFromMajorityCommittedSnapshot();
        if (!status.isOK()) {
            return status;
        }
    } else {
        invariant(readConcernLevel == repl::ReadConcernLevel::kLocalReadConcern);
    }

    DBDirectClient client(opCtx);
    Query fullQuery(query);
    if (!sort.isEmpty()) {
        fullQuery.sort(sort);
    }
    fullQuery.readPref(readPref.pref, BSONArray());

    try {
        std::unique_ptr<DBClientCursor> cursor =
            client.query(nss.ns().c_str(), fullQuery, limit.get_value_or(0));

        if (!cursor) {
            return {ErrorCodes::OperationFailed,
                    str::stream() << "Failed to establish a cursor for reading " << nss.ns()
                                  << " from local storage"};
        }

        std::vector<BSONObj> documentVector;
        while (cursor->more()) {
            BSONObj document = cursor->nextSafe().getOwned();
            documentVector.push_back(std::move(document));
        }

        return Shard::QueryResponse{std::move(documentVector),
                                    replCoord->getCurrentCommittedSnapshotOpTime()};
    } catch (const DBException& ex) {
        return ex.toStatus();
    }
}

Status ShardLocal::createIndexOnConfig(OperationContext* opCtx,
                                       const NamespaceString& ns,
                                       const BSONObj& keys,
                                       bool unique) {
    invariant(ns.db() == "config" || ns.db() == "admin");

    try {
        DBDirectClient client(opCtx);
        IndexSpec index;
        index.addKeys(keys);
        index.unique(unique);
        client.createIndex(ns.toString(), index);
    } catch (const DBException& e) {
        return e.toStatus();
    }

    return Status::OK();
}

}  // namespace mongo