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