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