includes/libs/rdbms/loadbalancer/ILoadBalancer.php

   1 <?php
   2 /**
   3  * Database load balancing interface
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Database
  22  */
  23 namespace Wikimedia\Rdbms;
  24
  25 use Exception;
  26 use InvalidArgumentException;
  27
  28 /**
  29  * Database cluster connection, tracking, load balancing, and transaction manager interface
  30  *
  31  * A "cluster" is considered to be one master database and zero or more replica databases.
  32  * Typically, the replica DBs replicate from the master asynchronously. The first node in the
  33  * "servers" configuration array is always considered the "master". However, this class can still
  34  * be used when all or some of the "replica" DBs are multi-master peers of the master or even
  35  * when all the DBs are non-replicating clones of each other holding read-only data. Thus, the
  36  * role of "master" is in some cases merely nominal.
  37  *
  38  * By default, each DB server uses DBO_DEFAULT for its 'flags' setting, unless explicitly set
  39  * otherwise in configuration. DBO_DEFAULT behavior depends on whether 'cliMode' is set:
  40  *   - In CLI mode, the flag has no effect with regards to LoadBalancer.
  41  *   - In non-CLI mode, the flag causes implicit transactions to be used; the first query on
  42  *     a database starts a transaction on that database. The transactions are meant to remain
  43  *     pending until either commitMasterChanges() or rollbackMasterChanges() is called. The
  44  *     application must have some point where it calls commitMasterChanges() near the end of
  45  *     the PHP request.
  46  * Every iteration of beginMasterChanges()/commitMasterChanges() is called a "transaction round".
  47  * Rounds are useful on the master DB connections because they make single-DB (and by and large
  48  * multi-DB) updates in web requests all-or-nothing. Also, transactions on replica DBs are useful
  49  * when REPEATABLE-READ or SERIALIZABLE isolation is used because all foriegn keys and constraints
  50  * hold across separate queries in the DB transaction since the data appears within a consistent
  51  * point-in-time snapshot.
  52  *
  53  * The typical caller will use LoadBalancer::getConnection( DB_* ) to yield a live database
  54  * connection handle. The choice of which DB server to use is based on pre-defined loads for
  55  * weighted random selection, adjustments thereof by LoadMonitor, and the amount of replication
  56  * lag on each DB server. Lag checks might cause problems in certain setups, so they should be
  57  * tuned in the server configuration maps as follows:
  58  *   - Master + N Replica(s): set 'max lag' to an appropriate threshold for avoiding any database
  59  *      lagged by this much or more. If all DBs are this lagged, then the load balancer considers
  60  *      the cluster to be read-only.
  61  *   - Galera Cluster: Seconds_Behind_Master will be 0, so there probably is nothing to tune.
  62  *      Note that lag is still possible depending on how wsrep-sync-wait is set server-side.
  63  *   - Read-only archive clones: set 'is static' in the server configuration maps. This will
  64  *      treat all such DBs as having 0 lag.
  65  *   - SQL load balancing proxy: any proxy should handle lag checks on its own, so the 'max lag'
  66  *      parameter should probably be set to INF in the server configuration maps. This will make
  67  *      the load balancer ignore whatever it detects as the lag of the logical replica is (which
  68  *      would probably just randomly bounce around).
  69  *
  70  * If using a SQL proxy service, it would probably be best to have two proxy hosts for the
  71  * load balancer to talk to. One would be the 'host' of the master server entry and another for
  72  * the (logical) replica server entry. The proxy could map the load balancer's "replica" DB to
  73  * any number of physical replica DBs.
  74  *
  75  * @since 1.28
  76  * @ingroup Database
  77  */
  78 interface ILoadBalancer {
  79         /** @var int Request a replica DB connection */
  80         const DB_REPLICA = -1;
  81         /** @var int Request a master DB connection */
  82         const DB_MASTER = -2;
  83
  84         /** @var string Domain specifier when no specific database needs to be selected */
  85         const DOMAIN_ANY = '';
  86
  87         /** @var int DB handle should have DBO_TRX disabled and the caller will leave it as such */
  88         const CONN_TRX_AUTO = 1;
  89
  90         /**
  91          * Construct a manager of IDatabase connection objects
  92          *
  93          * @param array $params Parameter map with keys:
  94          *  - servers : Required. Array of server info structures.
  95          *  - localDomain: A DatabaseDomain or domain ID string.
  96          *  - loadMonitor : Name of a class used to fetch server lag and load.
  97          *  - readOnlyReason : Reason the master DB is read-only if so [optional]
  98          *  - waitTimeout : Maximum time to wait for replicas for consistency [optional]
  99          *  - srvCache : BagOStuff object for server cache [optional]
 100          *  - wanCache : WANObjectCache object [optional]
 101          *  - chronologyProtector: ChronologyProtector object [optional]
 102          *  - hostname : The name of the current server [optional]
 103          *  - cliMode: Whether the execution context is a CLI script. [optional]
 104          *  - profiler : Class name or instance with profileIn()/profileOut() methods. [optional]
 105          *  - trxProfiler: TransactionProfiler instance. [optional]
 106          *  - replLogger: PSR-3 logger instance. [optional]
 107          *  - connLogger: PSR-3 logger instance. [optional]
 108          *  - queryLogger: PSR-3 logger instance. [optional]
 109          *  - perfLogger: PSR-3 logger instance. [optional]
 110          *  - errorLogger : Callback that takes an Exception and logs it. [optional]
 111          * @throws InvalidArgumentException
 112          */
 113         public function __construct( array $params );
 114
 115         /**
 116          * Get the index of the reader connection, which may be a replica DB
 117          *
 118          * This takes into account load ratios and lag times. It should
 119          * always return a consistent index during a given invocation.
 120          *
 121          * Side effect: opens connections to databases
 122          * @param string|bool $group Query group, or false for the generic reader
 123          * @param string|bool $domain Domain ID, or false for the current domain
 124          * @throws DBError
 125          * @return bool|int|string
 126          */
 127         public function getReaderIndex( $group = false, $domain = false );
 128
 129         /**
 130          * Set the master wait position
 131          *
 132          * If a DB_REPLICA connection has been opened already, then wait immediately.
 133          * Otherwise sets a variable telling it to wait if such a connection is opened.
 134          *
 135          * This only applies to connections to the generic replica DB for this request.
 136          * If a timeout happens when waiting, then getLaggedReplicaMode()/laggedReplicaUsed()
 137          * will return true.
 138          *
 139          * @param DBMasterPos|bool $pos Master position or false
 140          */
 141         public function waitFor( $pos );
 142
 143         /**
 144          * Set the master wait position and wait for a "generic" replica DB to catch up to it
 145          *
 146          * This can be used a faster proxy for waitForAll()
 147          *
 148          * @param DBMasterPos|bool $pos Master position or false
 149          * @param int $timeout Max seconds to wait; default is mWaitTimeout
 150          * @return bool Success (able to connect and no timeouts reached)
 151          */
 152         public function waitForOne( $pos, $timeout = null );
 153
 154         /**
 155          * Set the master wait position and wait for ALL replica DBs to catch up to it
 156          *
 157          * @param DBMasterPos|bool $pos Master position or false
 158          * @param int $timeout Max seconds to wait; default is mWaitTimeout
 159          * @return bool Success (able to connect and no timeouts reached)
 160          */
 161         public function waitForAll( $pos, $timeout = null );
 162
 163         /**
 164          * Get any open connection to a given server index, local or foreign
 165          *
 166          * @param int $i Server index or DB_MASTER/DB_REPLICA
 167          * @return Database|bool False if no such connection is open
 168          */
 169         public function getAnyOpenConnection( $i );
 170
 171         /**
 172          * Get a connection by index
 173          *
 174          * Avoid using CONN_TRX_AUTO with sqlite (e.g. check getServerType() first)
 175          *
 176          * @param int $i Server index or DB_MASTER/DB_REPLICA
 177          * @param array|string|bool $groups Query group(s), or false for the generic reader
 178          * @param string|bool $domain Domain ID, or false for the current domain
 179          * @param int $flags Bitfield of CONN_* class constants
 180          *
 181          * @throws DBError
 182          * @return Database
 183          */
 184         public function getConnection( $i, $groups = [], $domain = false, $flags = 0 );
 185
 186         /**
 187          * Mark a foreign connection as being available for reuse under a different DB domain
 188          *
 189          * This mechanism is reference-counted, and must be called the same number of times
 190          * as getConnection() to work.
 191          *
 192          * @param IDatabase $conn
 193          * @throws InvalidArgumentException
 194          */
 195         public function reuseConnection( $conn );
 196
 197         /**
 198          * Get a database connection handle reference
 199          *
 200          * The handle's methods simply wrap those of a Database handle
 201          *
 202          * Avoid using CONN_TRX_AUTO with sqlite (e.g. check getServerType() first)
 203          *
 204          * @see ILoadBalancer::getConnection() for parameter information
 205          *
 206          * @param int $i Server index or DB_MASTER/DB_REPLICA
 207          * @param array|string|bool $groups Query group(s), or false for the generic reader
 208          * @param string|bool $domain Domain ID, or false for the current domain
 209          * @param int $flags Bitfield of CONN_* class constants (e.g. CONN_TRX_AUTO)
 210          * @return DBConnRef
 211          */
 212         public function getConnectionRef( $i, $groups = [], $domain = false, $flags = 0 );
 213
 214         /**
 215          * Get a database connection handle reference without connecting yet
 216          *
 217          * The handle's methods simply wrap those of a Database handle
 218          *
 219          * Avoid using CONN_TRX_AUTO with sqlite (e.g. check getServerType() first)
 220          *
 221          * @see ILoadBalancer::getConnection() for parameter information
 222          *
 223          * @param int $i Server index or DB_MASTER/DB_REPLICA
 224          * @param array|string|bool $groups Query group(s), or false for the generic reader
 225          * @param string|bool $domain Domain ID, or false for the current domain
 226          * @param int $flags Bitfield of CONN_* class constants (e.g. CONN_TRX_AUTO)
 227          * @return DBConnRef
 228          */
 229         public function getLazyConnectionRef( $i, $groups = [], $domain = false, $flags = 0 );
 230
 231         /**
 232          * Get a maintenance database connection handle reference for migrations and schema changes
 233          *
 234          * The handle's methods simply wrap those of a Database handle
 235          *
 236          * Avoid using CONN_TRX_AUTO with sqlite (e.g. check getServerType() first)
 237          *
 238          * @see ILoadBalancer::getConnection() for parameter information
 239          *
 240          * @param int $db Server index or DB_MASTER/DB_REPLICA
 241          * @param array|string|bool $groups Query group(s), or false for the generic reader
 242          * @param string|bool $domain Domain ID, or false for the current domain
 243          * @param int $flags Bitfield of CONN_* class constants (e.g. CONN_TRX_AUTO)
 244          * @return MaintainableDBConnRef
 245          */
 246         public function getMaintenanceConnectionRef( $db, $groups = [], $domain = false, $flags = 0 );
 247
 248         /**
 249          * Open a connection to the server given by the specified index
 250          *
 251          * The index must be an actual index into the array. If a connection to the server is
 252          * already open and not considered an "in use" foreign connection, this simply returns it.
 253          *
 254          * Avoid using CONN_TRX_AUTO with sqlite (e.g. check getServerType() first)
 255          *
 256          * @note If disable() was called on this LoadBalancer, this method will throw a DBAccessError.
 257          *
 258          * @param int $i Server index (does not support DB_MASTER/DB_REPLICA)
 259          * @param string|bool $domain Domain ID, or false for the current domain
 260          * @param int $flags Bitfield of CONN_* class constants (e.g. CONN_TRX_AUTO)
 261          * @return Database|bool Returns false on errors
 262          * @throws DBAccessError
 263          */
 264         public function openConnection( $i, $domain = false, $flags = 0 );
 265
 266         /**
 267          * @return int
 268          */
 269         public function getWriterIndex();
 270
 271         /**
 272          * Returns true if the specified index is a valid server index
 273          *
 274          * @param int $i
 275          * @return bool
 276          */
 277         public function haveIndex( $i );
 278
 279         /**
 280          * Returns true if the specified index is valid and has non-zero load
 281          *
 282          * @param int $i
 283          * @return bool
 284          */
 285         public function isNonZeroLoad( $i );
 286
 287         /**
 288          * Get the number of defined servers (not the number of open connections)
 289          *
 290          * @return int
 291          */
 292         public function getServerCount();
 293
 294         /**
 295          * Get the host name or IP address of the server with the specified index
 296          *
 297          * @param int $i
 298          * @return string Readable name if available or IP/host otherwise
 299          */
 300         public function getServerName( $i );
 301
 302         /**
 303          * Get DB type of the server with the specified index
 304          *
 305          * @param int $i
 306          * @return string One of (mysql,postgres,sqlite,...) or "unknown" for bad indexes
 307          * @since 1.30
 308          */
 309         public function getServerType( $i );
 310
 311         /**
 312          * Return the server info structure for a given index, or false if the index is invalid.
 313          * @param int $i
 314          * @return array|bool
 315          *
 316          * @deprecated Since 1.30, no alternative
 317          */
 318         public function getServerInfo( $i );
 319
 320         /**
 321          * Sets the server info structure for the given index. Entry at index $i
 322          * is created if it doesn't exist
 323          * @param int $i
 324          * @param array $serverInfo
 325          *
 326          * @deprecated Since 1.30, construct new object
 327          */
 328         public function setServerInfo( $i, array $serverInfo );
 329
 330         /**
 331          * Get the current master position for chronology control purposes
 332          * @return DBMasterPos|bool Returns false if not applicable
 333          */
 334         public function getMasterPos();
 335
 336         /**
 337          * Disable this load balancer. All connections are closed, and any attempt to
 338          * open a new connection will result in a DBAccessError.
 339          */
 340         public function disable();
 341
 342         /**
 343          * Close all open connections
 344          */
 345         public function closeAll();
 346
 347         /**
 348          * Close a connection
 349          *
 350          * Using this function makes sure the LoadBalancer knows the connection is closed.
 351          * If you use $conn->close() directly, the load balancer won't update its state.
 352          *
 353          * @param IDatabase $conn
 354          */
 355         public function closeConnection( IDatabase $conn );
 356
 357         /**
 358          * Commit transactions on all open connections
 359          * @param string $fname Caller name
 360          * @throws DBExpectedError
 361          */
 362         public function commitAll( $fname = __METHOD__ );
 363
 364         /**
 365          * Perform all pre-commit callbacks that remain part of the atomic transactions
 366          * and disable any post-commit callbacks until runMasterPostTrxCallbacks()
 367          *
 368          * Use this only for mutli-database commits
 369          */
 370         public function finalizeMasterChanges();
 371
 372         /**
 373          * Perform all pre-commit checks for things like replication safety
 374          *
 375          * Use this only for mutli-database commits
 376          *
 377          * @param array $options Includes:
 378          *   - maxWriteDuration : max write query duration time in seconds
 379          * @throws DBTransactionError
 380          */
 381         public function approveMasterChanges( array $options );
 382
 383         /**
 384          * Flush any master transaction snapshots and set DBO_TRX (if DBO_DEFAULT is set)
 385          *
 386          * The DBO_TRX setting will be reverted to the default in each of these methods:
 387          *   - commitMasterChanges()
 388          *   - rollbackMasterChanges()
 389          *   - commitAll()
 390          * This allows for custom transaction rounds from any outer transaction scope.
 391          *
 392          * @param string $fname
 393          * @throws DBExpectedError
 394          */
 395         public function beginMasterChanges( $fname = __METHOD__ );
 396
 397         /**
 398          * Issue COMMIT on all master connections where writes where done
 399          * @param string $fname Caller name
 400          * @throws DBExpectedError
 401          */
 402         public function commitMasterChanges( $fname = __METHOD__ );
 403
 404         /**
 405          * Issue all pending post-COMMIT/ROLLBACK callbacks
 406          *
 407          * Use this only for mutli-database commits
 408          *
 409          * @param int $type IDatabase::TRIGGER_* constant
 410          * @return Exception|null The first exception or null if there were none
 411          */
 412         public function runMasterPostTrxCallbacks( $type );
 413
 414         /**
 415          * Issue ROLLBACK only on master, only if queries were done on connection
 416          * @param string $fname Caller name
 417          * @throws DBExpectedError
 418          */
 419         public function rollbackMasterChanges( $fname = __METHOD__ );
 420
 421         /**
 422          * Suppress all pending post-COMMIT/ROLLBACK callbacks
 423          *
 424          * Use this only for mutli-database commits
 425          *
 426          * @return Exception|null The first exception or null if there were none
 427          */
 428         public function suppressTransactionEndCallbacks();
 429
 430         /**
 431          * Commit all replica DB transactions so as to flush any REPEATABLE-READ or SSI snapshot
 432          *
 433          * @param string $fname Caller name
 434          */
 435         public function flushReplicaSnapshots( $fname = __METHOD__ );
 436
 437         /**
 438          * @return bool Whether a master connection is already open
 439          */
 440         public function hasMasterConnection();
 441
 442         /**
 443          * Determine if there are pending changes in a transaction by this thread
 444          * @return bool
 445          */
 446         public function hasMasterChanges();
 447
 448         /**
 449          * Get the timestamp of the latest write query done by this thread
 450          * @return float|bool UNIX timestamp or false
 451          */
 452         public function lastMasterChangeTimestamp();
 453
 454         /**
 455          * Check if this load balancer object had any recent or still
 456          * pending writes issued against it by this PHP thread
 457          *
 458          * @param float $age How many seconds ago is "recent" [defaults to mWaitTimeout]
 459          * @return bool
 460          */
 461         public function hasOrMadeRecentMasterChanges( $age = null );
 462
 463         /**
 464          * Get the list of callers that have pending master changes
 465          *
 466          * @return string[] List of method names
 467          */
 468         public function pendingMasterChangeCallers();
 469
 470         /**
 471          * @note This method will trigger a DB connection if not yet done
 472          * @param string|bool $domain Domain ID, or false for the current domain
 473          * @return bool Whether the database for generic connections this request is highly "lagged"
 474          */
 475         public function getLaggedReplicaMode( $domain = false );
 476
 477         /**
 478          * Checks whether the database for generic connections this request was both:
 479          *   - a) Already choosen due to a prior connection attempt
 480          *   - b) Considered highly "lagged"
 481          *
 482          * @note This method will never cause a new DB connection
 483          * @return bool
 484          */
 485         public function laggedReplicaUsed();
 486
 487         /**
 488          * @note This method may trigger a DB connection if not yet done
 489          * @param string|bool $domain Domain ID, or false for the current domain
 490          * @param IDatabase|null $conn DB master connection; used to avoid loops [optional]
 491          * @return string|bool Reason the master is read-only or false if it is not
 492          */
 493         public function getReadOnlyReason( $domain = false, IDatabase $conn = null );
 494
 495         /**
 496          * Disables/enables lag checks
 497          * @param null|bool $mode
 498          * @return bool
 499          */
 500         public function allowLagged( $mode = null );
 501
 502         /**
 503          * @return bool
 504          */
 505         public function pingAll();
 506
 507         /**
 508          * Call a function with each open connection object
 509          * @param callable $callback
 510          * @param array $params
 511          */
 512         public function forEachOpenConnection( $callback, array $params = [] );
 513
 514         /**
 515          * Call a function with each open connection object to a master
 516          * @param callable $callback
 517          * @param array $params
 518          */
 519         public function forEachOpenMasterConnection( $callback, array $params = [] );
 520
 521         /**
 522          * Call a function with each open replica DB connection object
 523          * @param callable $callback
 524          * @param array $params
 525          */
 526         public function forEachOpenReplicaConnection( $callback, array $params = [] );
 527
 528         /**
 529          * Get the hostname and lag time of the most-lagged replica DB
 530          *
 531          * This is useful for maintenance scripts that need to throttle their updates.
 532          * May attempt to open connections to replica DBs on the default DB. If there is
 533          * no lag, the maximum lag will be reported as -1.
 534          *
 535          * @param bool|string $domain Domain ID, or false for the default database
 536          * @return array ( host, max lag, index of max lagged host )
 537          */
 538         public function getMaxLag( $domain = false );
 539
 540         /**
 541          * Get an estimate of replication lag (in seconds) for each server
 542          *
 543          * Results are cached for a short time in memcached/process cache
 544          *
 545          * Values may be "false" if replication is too broken to estimate
 546          *
 547          * @param string|bool $domain
 548          * @return int[] Map of (server index => float|int|bool)
 549          */
 550         public function getLagTimes( $domain = false );
 551
 552         /**
 553          * Get the lag in seconds for a given connection, or zero if this load
 554          * balancer does not have replication enabled.
 555          *
 556          * This should be used in preference to Database::getLag() in cases where
 557          * replication may not be in use, since there is no way to determine if
 558          * replication is in use at the connection level without running
 559          * potentially restricted queries such as SHOW SLAVE STATUS. Using this
 560          * function instead of Database::getLag() avoids a fatal error in this
 561          * case on many installations.
 562          *
 563          * @param IDatabase $conn
 564          * @return int|bool Returns false on error
 565          */
 566         public function safeGetLag( IDatabase $conn );
 567
 568         /**
 569          * Wait for a replica DB to reach a specified master position
 570          *
 571          * This will connect to the master to get an accurate position if $pos is not given
 572          *
 573          * @param IDatabase $conn Replica DB
 574          * @param DBMasterPos|bool $pos Master position; default: current position
 575          * @param int $timeout Timeout in seconds [optional]
 576          * @return bool Success
 577          */
 578         public function safeWaitForMasterPos( IDatabase $conn, $pos = false, $timeout = 10 );
 579
 580         /**
 581          * Set a callback via IDatabase::setTransactionListener() on
 582          * all current and future master connections of this load balancer
 583          *
 584          * @param string $name Callback name
 585          * @param callable|null $callback
 586          */
 587         public function setTransactionListener( $name, callable $callback = null );
 588
 589         /**
 590          * Set a new table prefix for the existing local domain ID for testing
 591          *
 592          * @param string $prefix
 593          */
 594         public function setDomainPrefix( $prefix );
 595
 596         /**
 597          * Make certain table names use their own database, schema, and table prefix
 598          * when passed into SQL queries pre-escaped and without a qualified database name
 599          *
 600          * For example, "user" can be converted to "myschema.mydbname.user" for convenience.
 601          * Appearances like `user`, somedb.user, somedb.someschema.user will used literally.
 602          *
 603          * Calling this twice will completely clear any old table aliases. Also, note that
 604          * callers are responsible for making sure the schemas and databases actually exist.
 605          *
 606          * @param array[] $aliases Map of (table => (dbname, schema, prefix) map)
 607          */
 608         public function setTableAliases( array $aliases );
 609 }