3 * Database load balancing interface
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.
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.
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
23 namespace Wikimedia\Rdbms
;
26 use InvalidArgumentException
;
29 * Database cluster connection, tracking, load balancing, and transaction manager interface
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.
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
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.
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).
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.
78 interface ILoadBalancer
{
79 /** @var int Request a replica DB connection */
80 const DB_REPLICA
= -1;
81 /** @var int Request a master DB connection */
84 /** @var string Domain specifier when no specific database needs to be selected */
85 const DOMAIN_ANY
= '';
87 /** @var int DB handle should have DBO_TRX disabled and the caller will leave it as such */
88 const CONN_TRX_AUTOCOMMIT
= 1;
89 /** @var int Alias for CONN_TRX_AUTOCOMMIT for b/c; deprecated since 1.31 */
90 const CONN_TRX_AUTO
= 1;
92 /** @var string Manager of ILoadBalancer instances is running post-commit callbacks */
93 const STAGE_POSTCOMMIT_CALLBACKS
= 'stage-postcommit-callbacks';
94 /** @var string Manager of ILoadBalancer instances is running post-rollback callbacks */
95 const STAGE_POSTROLLBACK_CALLBACKS
= 'stage-postrollback-callbacks';
98 * Construct a manager of IDatabase connection objects
100 * @param array $params Parameter map with keys:
101 * - servers : Required. Array of server info structures.
102 * - localDomain: A DatabaseDomain or domain ID string.
103 * - loadMonitor : Name of a class used to fetch server lag and load.
104 * - readOnlyReason : Reason the master DB is read-only if so [optional]
105 * - waitTimeout : Maximum time to wait for replicas for consistency [optional]
106 * - maxLag: Avoid replica DB servers with more lag than this [optional]
107 * - srvCache : BagOStuff object for server cache [optional]
108 * - wanCache : WANObjectCache object [optional]
109 * - chronologyCallback: Callback to run before the first connection attempt [optional]
110 * - hostname : The name of the current server [optional]
111 * - cliMode: Whether the execution context is a CLI script. [optional]
112 * - profiler : Class name or instance with profileIn()/profileOut() methods. [optional]
113 * - trxProfiler: TransactionProfiler instance. [optional]
114 * - replLogger: PSR-3 logger instance. [optional]
115 * - connLogger: PSR-3 logger instance. [optional]
116 * - queryLogger: PSR-3 logger instance. [optional]
117 * - perfLogger: PSR-3 logger instance. [optional]
118 * - errorLogger : Callback that takes an Exception and logs it. [optional]
119 * - deprecationLogger: Callback to log a deprecation warning. [optional]
120 * - roundStage: STAGE_POSTCOMMIT_* class constant; for internal use [optional]
121 * @throws InvalidArgumentException
123 public function __construct( array $params );
126 * Get the index of the reader connection, which may be a replica DB
128 * This takes into account load ratios and lag times. It should
129 * always return a consistent index during a given invocation.
131 * Side effect: opens connections to databases
132 * @param string|bool $group Query group, or false for the generic reader
133 * @param string|bool $domain Domain ID, or false for the current domain
135 * @return bool|int|string
137 public function getReaderIndex( $group = false, $domain = false );
140 * Set the master wait position
142 * If a DB_REPLICA connection has been opened already, then wait immediately.
143 * Otherwise sets a variable telling it to wait if such a connection is opened.
145 * This only applies to connections to the generic replica DB for this request.
146 * If a timeout happens when waiting, then getLaggedReplicaMode()/laggedReplicaUsed()
149 * @param DBMasterPos|bool $pos Master position or false
151 public function waitFor( $pos );
154 * Set the master wait position and wait for a "generic" replica DB to catch up to it
156 * This can be used a faster proxy for waitForAll()
158 * @param DBMasterPos|bool $pos Master position or false
159 * @param int $timeout Max seconds to wait; default is mWaitTimeout
160 * @return bool Success (able to connect and no timeouts reached)
162 public function waitForOne( $pos, $timeout = null );
165 * Set the master wait position and wait for ALL replica DBs to catch up to it
167 * @param DBMasterPos|bool $pos Master position or false
168 * @param int $timeout Max seconds to wait; default is mWaitTimeout
169 * @return bool Success (able to connect and no timeouts reached)
171 public function waitForAll( $pos, $timeout = null );
174 * Get any open connection to a given server index, local or foreign
176 * Use CONN_TRX_AUTOCOMMIT to only look for connections opened with that flag
178 * @param int $i Server index or DB_MASTER/DB_REPLICA
179 * @param int $flags Bitfield of CONN_* class constants
180 * @return Database|bool False if no such connection is open
182 public function getAnyOpenConnection( $i, $flags = 0 );
185 * Get a connection handle by server index
187 * The CONN_TRX_AUTOCOMMIT flag is ignored for databases with ATTR_DB_LEVEL_LOCKING
188 * (e.g. sqlite) in order to avoid deadlocks. ILoadBalancer::getServerAttributes()
189 * can be used to check such flags beforehand.
191 * If the caller uses $domain or sets CONN_TRX_AUTOCOMMIT in $flags, then it must also
192 * call ILoadBalancer::reuseConnection() on the handle when finished using it.
193 * In all other cases, this is not necessary, though not harmful either.
195 * @param int $i Server index (overrides $groups) or DB_MASTER/DB_REPLICA
196 * @param array|string|bool $groups Query group(s), or false for the generic reader
197 * @param string|bool $domain Domain ID, or false for the current domain
198 * @param int $flags Bitfield of CONN_* class constants
200 * @note This method throws DBAccessError if ILoadBalancer::disable() was called
205 public function getConnection( $i, $groups = [], $domain = false, $flags = 0 );
208 * Mark a foreign connection as being available for reuse under a different DB domain
210 * This mechanism is reference-counted, and must be called the same number of times
211 * as getConnection() to work.
213 * @param IDatabase $conn
214 * @throws InvalidArgumentException
216 public function reuseConnection( IDatabase
$conn );
219 * Get a database connection handle reference
221 * The handle's methods simply wrap those of a Database handle
223 * The CONN_TRX_AUTOCOMMIT flag is ignored for databases with ATTR_DB_LEVEL_LOCKING
224 * (e.g. sqlite) in order to avoid deadlocks. ILoadBalancer::getServerAttributes()
225 * can be used to check such flags beforehand.
227 * @see ILoadBalancer::getConnection() for parameter information
229 * @param int $i Server index or DB_MASTER/DB_REPLICA
230 * @param array|string|bool $groups Query group(s), or false for the generic reader
231 * @param string|bool $domain Domain ID, or false for the current domain
232 * @param int $flags Bitfield of CONN_* class constants (e.g. CONN_TRX_AUTOCOMMIT)
235 public function getConnectionRef( $i, $groups = [], $domain = false, $flags = 0 );
238 * Get a database connection handle reference without connecting yet
240 * The handle's methods simply wrap those of a Database handle
242 * The CONN_TRX_AUTOCOMMIT flag is ignored for databases with ATTR_DB_LEVEL_LOCKING
243 * (e.g. sqlite) in order to avoid deadlocks. ILoadBalancer::getServerAttributes()
244 * can be used to check such flags beforehand.
246 * @see ILoadBalancer::getConnection() for parameter information
248 * @param int $i Server index or DB_MASTER/DB_REPLICA
249 * @param array|string|bool $groups Query group(s), or false for the generic reader
250 * @param string|bool $domain Domain ID, or false for the current domain
251 * @param int $flags Bitfield of CONN_* class constants (e.g. CONN_TRX_AUTOCOMMIT)
254 public function getLazyConnectionRef( $i, $groups = [], $domain = false, $flags = 0 );
257 * Get a maintenance database connection handle reference for migrations and schema changes
259 * The handle's methods simply wrap those of a Database handle
261 * The CONN_TRX_AUTOCOMMIT flag is ignored for databases with ATTR_DB_LEVEL_LOCKING
262 * (e.g. sqlite) in order to avoid deadlocks. ILoadBalancer::getServerAttributes()
263 * can be used to check such flags beforehand.
265 * @see ILoadBalancer::getConnection() for parameter information
267 * @param int $db Server index or DB_MASTER/DB_REPLICA
268 * @param array|string|bool $groups Query group(s), or false for the generic reader
269 * @param string|bool $domain Domain ID, or false for the current domain
270 * @param int $flags Bitfield of CONN_* class constants (e.g. CONN_TRX_AUTOCOMMIT)
271 * @return MaintainableDBConnRef
273 public function getMaintenanceConnectionRef( $db, $groups = [], $domain = false, $flags = 0 );
276 * Open a connection to the server given by the specified index
278 * The index must be an actual index into the array. If a connection to the server is
279 * already open and not considered an "in use" foreign connection, this simply returns it.
281 * Avoid using CONN_TRX_AUTOCOMMIT for databases with ATTR_DB_LEVEL_LOCKING (e.g. sqlite) in
282 * order to avoid deadlocks. ILoadBalancer::getServerAttributes() can be used to check
283 * such flags beforehand.
285 * If the caller uses $domain or sets CONN_TRX_AUTOCOMMIT in $flags, then it must also
286 * call ILoadBalancer::reuseConnection() on the handle when finished using it.
287 * In all other cases, this is not necessary, though not harmful either.
289 * @note This method throws DBAccessError if ILoadBalancer::disable() was called
291 * @param int $i Server index (does not support DB_MASTER/DB_REPLICA)
292 * @param string|bool $domain Domain ID, or false for the current domain
293 * @param int $flags Bitfield of CONN_* class constants (e.g. CONN_TRX_AUTOCOMMIT)
294 * @return Database|bool Returns false on errors
295 * @throws DBAccessError
297 public function openConnection( $i, $domain = false, $flags = 0 );
302 public function getWriterIndex();
305 * Returns true if the specified index is a valid server index
310 public function haveIndex( $i );
313 * Returns true if the specified index is valid and has non-zero load
318 public function isNonZeroLoad( $i );
321 * Get the number of defined servers (not the number of open connections)
325 public function getServerCount();
328 * Get the host name or IP address of the server with the specified index
331 * @return string Readable name if available or IP/host otherwise
333 public function getServerName( $i );
336 * Get DB type of the server with the specified index
339 * @return string One of (mysql,postgres,sqlite,...) or "unknown" for bad indexes
342 public function getServerType( $i );
345 * @param int $i Server index
346 * @return array (Database::ATTRIBUTE_* constant => value) for all such constants
349 public function getServerAttributes( $i );
352 * Get the current master position for chronology control purposes
353 * @return DBMasterPos|bool Returns false if not applicable
355 public function getMasterPos();
358 * Disable this load balancer. All connections are closed, and any attempt to
359 * open a new connection will result in a DBAccessError.
361 public function disable();
364 * Close all open connections
366 public function closeAll();
371 * Using this function makes sure the LoadBalancer knows the connection is closed.
372 * If you use $conn->close() directly, the load balancer won't update its state.
374 * @param IDatabase $conn
376 public function closeConnection( IDatabase
$conn );
379 * Commit transactions on all open connections
380 * @param string $fname Caller name
381 * @throws DBExpectedError
383 public function commitAll( $fname = __METHOD__
);
386 * Run pre-commit callbacks and defer execution of post-commit callbacks
388 * Use this only for mutli-database commits
390 * @return int Number of pre-commit callbacks run (since 1.32)
392 public function finalizeMasterChanges();
395 * Perform all pre-commit checks for things like replication safety
397 * Use this only for mutli-database commits
399 * @param array $options Includes:
400 * - maxWriteDuration : max write query duration time in seconds
401 * @throws DBTransactionError
403 public function approveMasterChanges( array $options );
406 * Flush any master transaction snapshots and set DBO_TRX (if DBO_DEFAULT is set)
408 * The DBO_TRX setting will be reverted to the default in each of these methods:
409 * - commitMasterChanges()
410 * - rollbackMasterChanges()
412 * This allows for custom transaction rounds from any outer transaction scope.
414 * @param string $fname
415 * @throws DBExpectedError
417 public function beginMasterChanges( $fname = __METHOD__
);
420 * Issue COMMIT on all master connections where writes where done
421 * @param string $fname Caller name
422 * @throws DBExpectedError
424 public function commitMasterChanges( $fname = __METHOD__
);
427 * Consume and run all pending post-COMMIT/ROLLBACK callbacks and commit dangling transactions
429 * @return Exception|null The first exception or null if there were none
431 public function runMasterTransactionIdleCallbacks();
434 * Run all recurring post-COMMIT/ROLLBACK listener callbacks
436 * @return Exception|null The first exception or null if there were none
438 public function runMasterTransactionListenerCallbacks();
441 * Issue ROLLBACK only on master, only if queries were done on connection
442 * @param string $fname Caller name
443 * @throws DBExpectedError
445 public function rollbackMasterChanges( $fname = __METHOD__
);
448 * Commit all replica DB transactions so as to flush any REPEATABLE-READ or SSI snapshots
450 * @param string $fname Caller name
452 public function flushReplicaSnapshots( $fname = __METHOD__
);
455 * Commit all master DB transactions so as to flush any REPEATABLE-READ or SSI snapshots
457 * An error will be thrown if a connection has pending writes or callbacks
459 * @param string $fname Caller name
461 public function flushMasterSnapshots( $fname = __METHOD__
);
464 * @return bool Whether a master connection is already open
466 public function hasMasterConnection();
469 * Whether there are pending changes or callbacks in a transaction by this thread
472 public function hasMasterChanges();
475 * Get the timestamp of the latest write query done by this thread
476 * @return float|bool UNIX timestamp or false
478 public function lastMasterChangeTimestamp();
481 * Check if this load balancer object had any recent or still
482 * pending writes issued against it by this PHP thread
484 * @param float $age How many seconds ago is "recent" [defaults to mWaitTimeout]
487 public function hasOrMadeRecentMasterChanges( $age = null );
490 * Get the list of callers that have pending master changes
492 * @return string[] List of method names
494 public function pendingMasterChangeCallers();
497 * @note This method will trigger a DB connection if not yet done
498 * @param string|bool $domain Domain ID, or false for the current domain
499 * @return bool Whether the database for generic connections this request is highly "lagged"
501 public function getLaggedReplicaMode( $domain = false );
504 * Checks whether the database for generic connections this request was both:
505 * - a) Already choosen due to a prior connection attempt
506 * - b) Considered highly "lagged"
508 * @note This method will never cause a new DB connection
511 public function laggedReplicaUsed();
514 * @note This method may trigger a DB connection if not yet done
515 * @param string|bool $domain Domain ID, or false for the current domain
516 * @param IDatabase|null $conn DB master connection; used to avoid loops [optional]
517 * @return string|bool Reason the master is read-only or false if it is not
519 public function getReadOnlyReason( $domain = false, IDatabase
$conn = null );
522 * Disables/enables lag checks
523 * @param null|bool $mode
526 public function allowLagged( $mode = null );
531 public function pingAll();
534 * Call a function with each open connection object
535 * @param callable $callback
536 * @param array $params
538 public function forEachOpenConnection( $callback, array $params = [] );
541 * Call a function with each open connection object to a master
542 * @param callable $callback
543 * @param array $params
545 public function forEachOpenMasterConnection( $callback, array $params = [] );
548 * Call a function with each open replica DB connection object
549 * @param callable $callback
550 * @param array $params
552 public function forEachOpenReplicaConnection( $callback, array $params = [] );
555 * Get the hostname and lag time of the most-lagged replica DB
557 * This is useful for maintenance scripts that need to throttle their updates.
558 * May attempt to open connections to replica DBs on the default DB. If there is
559 * no lag, the maximum lag will be reported as -1.
561 * @param bool|string $domain Domain ID, or false for the default database
562 * @return array ( host, max lag, index of max lagged host )
564 public function getMaxLag( $domain = false );
567 * Get an estimate of replication lag (in seconds) for each server
569 * Results are cached for a short time in memcached/process cache
571 * Values may be "false" if replication is too broken to estimate
573 * @param string|bool $domain
574 * @return int[] Map of (server index => float|int|bool)
576 public function getLagTimes( $domain = false );
579 * Get the lag in seconds for a given connection, or zero if this load
580 * balancer does not have replication enabled.
582 * This should be used in preference to Database::getLag() in cases where
583 * replication may not be in use, since there is no way to determine if
584 * replication is in use at the connection level without running
585 * potentially restricted queries such as SHOW SLAVE STATUS. Using this
586 * function instead of Database::getLag() avoids a fatal error in this
587 * case on many installations.
589 * @param IDatabase $conn
590 * @return int|bool Returns false on error
592 public function safeGetLag( IDatabase
$conn );
595 * Wait for a replica DB to reach a specified master position
597 * This will connect to the master to get an accurate position if $pos is not given
599 * @param IDatabase $conn Replica DB
600 * @param DBMasterPos|bool $pos Master position; default: current position
601 * @param int $timeout Timeout in seconds [optional]
602 * @return bool Success
604 public function safeWaitForMasterPos( IDatabase
$conn, $pos = false, $timeout = 10 );
607 * Set a callback via IDatabase::setTransactionListener() on
608 * all current and future master connections of this load balancer
610 * @param string $name Callback name
611 * @param callable|null $callback
613 public function setTransactionListener( $name, callable
$callback = null );
616 * Set a new table prefix for the existing local domain ID for testing
618 * @param string $prefix
620 public function setDomainPrefix( $prefix );
623 * Make certain table names use their own database, schema, and table prefix
624 * when passed into SQL queries pre-escaped and without a qualified database name
626 * For example, "user" can be converted to "myschema.mydbname.user" for convenience.
627 * Appearances like `user`, somedb.user, somedb.someschema.user will used literally.
629 * Calling this twice will completely clear any old table aliases. Also, note that
630 * callers are responsible for making sure the schemas and databases actually exist.
632 * @param array[] $aliases Map of (table => (dbname, schema, prefix) map)
634 public function setTableAliases( array $aliases );
637 * Convert certain index names to alternative names before querying the DB
639 * Note that this applies to indexes regardless of the table they belong to.
641 * This can be employed when an index was renamed X => Y in code, but the new Y-named
642 * indexes were not yet built on all DBs. After all the Y-named ones are added by the DBA,
643 * the aliases can be removed, and then the old X-named indexes dropped.
645 * @param string[] $aliases
649 public function setIndexAliases( array $aliases );