15.4.2.3 CHANGE REPLICATION SOURCE TO Statement (original) (raw)

15.4.2.3 CHANGE REPLICATION SOURCE TO Statement

CHANGE REPLICATION SOURCE TO option [, option] ... [ channel_option ]

option: {
    SOURCE_BIND = 'interface_name'
  | SOURCE_HOST = 'host_name'
  | SOURCE_USER = 'user_name'
  | SOURCE_PASSWORD = 'password'
  | SOURCE_PORT = port_num
  | PRIVILEGE_CHECKS_USER = {NULL | 'account'}
  | REQUIRE_ROW_FORMAT = {0|1}
  | REQUIRE_TABLE_PRIMARY_KEY_CHECK = {STREAM | ON | OFF | GENERATE}
  | ASSIGN_GTIDS_TO_ANONYMOUS_TRANSACTIONS = {OFF | LOCAL | uuid}
  | SOURCE_LOG_FILE = 'source_log_name'
  | SOURCE_LOG_POS = source_log_pos
  | SOURCE_AUTO_POSITION = {0|1}
  | RELAY_LOG_FILE = 'relay_log_name'
  | RELAY_LOG_POS = relay_log_pos
  | SOURCE_HEARTBEAT_PERIOD = interval
  | SOURCE_CONNECT_RETRY = interval
  | SOURCE_RETRY_COUNT = count
  | SOURCE_CONNECTION_AUTO_FAILOVER = {0|1}
  | SOURCE_DELAY = interval
  | SOURCE_COMPRESSION_ALGORITHMS = 'algorithm[,algorithm][,algorithm]'
  | SOURCE_ZSTD_COMPRESSION_LEVEL = level
  | SOURCE_SSL = {0|1}
  | SOURCE_SSL_CA = 'ca_file_name'
  | SOURCE_SSL_CAPATH = 'ca_directory_name'
  | SOURCE_SSL_CERT = 'cert_file_name'
  | SOURCE_SSL_CRL = 'crl_file_name'
  | SOURCE_SSL_CRLPATH = 'crl_directory_name'
  | SOURCE_SSL_KEY = 'key_file_name'
  | SOURCE_SSL_CIPHER = 'cipher_list'
  | SOURCE_SSL_VERIFY_SERVER_CERT = {0|1}
  | SOURCE_TLS_VERSION = 'protocol_list'
  | SOURCE_TLS_CIPHERSUITES = 'ciphersuite_list'
  | SOURCE_PUBLIC_KEY_PATH = 'key_file_name'
  | GET_SOURCE_PUBLIC_KEY = {0|1}
  | NETWORK_NAMESPACE = 'namespace'
  | IGNORE_SERVER_IDS = (server_id_list),
  | GTID_ONLY = {0|1}
}

channel_option:
    FOR CHANNEL channel

server_id_list:
    [server_id [, server_id] ... ]

CHANGE REPLICATION SOURCE TO changes the parameters that the replica server uses for connecting to the source and reading data from the source. It also updates the contents of the replication metadata repositories (see Section 19.2.4, “Relay Log and Replication Metadata Repositories”). In MySQL 8.0.23 and later, use CHANGE REPLICATION SOURCE TO in place of the deprecatedCHANGE MASTER TO statement.

CHANGE REPLICATION SOURCE TO requires theREPLICATION_SLAVE_ADMIN privilege (or the deprecated SUPER privilege).

Options that you do not specify on a CHANGE REPLICATION SOURCE TO statement retain their value, except as indicated in the following discussion. In most cases, there is therefore no need to specify options that do not change.

Values used for SOURCE_HOST and otherCHANGE REPLICATION SOURCE TO options are checked for linefeed (\n or0x0A) characters. The presence of such characters in these values causes the statement to fail with an error.

The optional FOR CHANNEL_`channel`_ clause lets you name which replication channel the statement applies to. Providing a FOR CHANNEL_`channel`_ clause applies theCHANGE REPLICATION SOURCE TO statement to a specific replication channel, and is used to add a new channel or modify an existing channel. For example, to add a new channel called channel2:

CHANGE REPLICATION SOURCE TO SOURCE_HOST=host1, SOURCE_PORT=3002 FOR CHANNEL 'channel2';

If no clause is named and no extra channels exist, aCHANGE REPLICATION SOURCE TO statement applies to the default channel, whose name is the empty string (""). When you have set up multiple replication channels, every CHANGE REPLICATION SOURCE TO statement must name a channel using theFOR CHANNEL_`channel`_ clause. SeeSection 19.2.2, “Replication Channels” for more information.

For some of the options of the CHANGE REPLICATION SOURCE TO statement, you must issue aSTOP REPLICA statement prior to issuing a CHANGE REPLICATION SOURCE TO statement (and a START REPLICA statement afterwards). Sometimes, you only need to stop the replication SQL (applier) thread or the replication I/O (receiver) thread, not both:

• When the applier thread is stopped, you can executeCHANGE REPLICATION SOURCE TO using any combination that is otherwise allowed ofRELAY_LOG_FILE,RELAY_LOG_POS, andSOURCE_DELAY options, even if the replication receiver thread is running. No other options may be used with this statement when the receiver thread is running.
• When the receiver thread is stopped, you can executeCHANGE REPLICATION SOURCE TO using any of the options for this statement (in any allowed combination) except RELAY_LOG_FILE,RELAY_LOG_POS,SOURCE_DELAY, orSOURCE_AUTO_POSITION = 1 even when the applier thread is running.
• Both the receiver thread and the applier thread must be stopped before issuing a CHANGE REPLICATION SOURCE TO statement that employsSOURCE_AUTO_POSITION = 1,GTID_ONLY = 1, orASSIGN_GTIDS_TO_ANONYMOUS_TRANSACTIONS.

You can check the current state of the replication applier thread and replication receiver thread usingSHOW REPLICA STATUS. Note that the Group Replication applier channel (group_replication_applier) has no receiver thread, only an applier thread.

CHANGE REPLICATION SOURCE TO statements have a number of side-effects and interactions that you should be aware of beforehand:

• CHANGE REPLICATION SOURCE TO causes an implicit commit of an ongoing transaction. SeeSection 15.3.3, “Statements That Cause an Implicit Commit”.
• CHANGE REPLICATION SOURCE TO causes the previous values forSOURCE_HOST,SOURCE_PORT,SOURCE_LOG_FILE, andSOURCE_LOG_POS to be written to the error log, along with other information about the replica's state prior to execution.
• If you are using statement-based replication and temporary tables, it is possible for a CHANGE REPLICATION SOURCE TO statement following aSTOP REPLICA statement to leave behind temporary tables on the replica. A warning (ER_WARN_OPEN_TEMP_TABLES_MUST_BE_ZERO) is issued whenever this occurs. You can avoid this in such cases by making sure that the value of theReplica_open_temp_tables or Slave_open_temp_tables system status variable is equal to 0 prior to executing such a CHANGE REPLICATION SOURCE TO statement.
• When using a multithreaded replica (replica_parallel_workers > 0), stopping the replica can cause gaps in the sequence of transactions that have been executed from the relay log, regardless of whether the replica was stopped intentionally or otherwise. When such gaps exist, issuingCHANGE REPLICATION SOURCE TO fails. The solution in this situation is to issueSTART REPLICA UNTIL SQL_AFTER_MTS_GAPS which ensures that the gaps are closed. From MySQL 8.0.26, the process of checking for gaps in the sequence of transactions is skipped entirely when GTID-based replication and GTID auto-positioning are in use, because gaps in transactions can be resolved using GTID auto-positioning. In that situation,CHANGE REPLICATION SOURCE TO can still be used.

The following options are available forCHANGE REPLICATION SOURCE TO statements:

• ASSIGN_GTIDS_TO_ANONYMOUS_TRANSACTIONS = {OFF | LOCAL | uuid}
  Makes the replication channel assign a GTID to replicated transactions that do not have one, enabling replication from a source that does not use GTID-based replication, to a replica that does. For a multi-source replica, you can have a mix of channels that useASSIGN_GTIDS_TO_ANONYMOUS_TRANSACTIONS, and channels that do not. The default isOFF, meaning that the feature is not used.
  LOCAL assigns a GTID including the replica's own UUID (theserver_uuid setting)._`uuid`_ assigns a GTID including the specified UUID, such as theserver_uuid setting for the replication source server. Using a nonlocal UUID lets you differentiate between transactions that originated on the replica and transactions that originated on the source, and for a multi-source replica, between transactions that originated on different sources. The UUID you choose only has significance for the replica's own use. If any of the transactions sent by the source do have a GTID already, that GTID is retained.
  Channels specific to Group Replication cannot useASSIGN_GTIDS_TO_ANONYMOUS_TRANSACTIONS, but an asynchronous replication channel for another source on a server instance that is a Group Replication group member can do so. In that case, do not specify the Group Replication group name as the UUID for creating the GTIDs.
  To setASSIGN_GTIDS_TO_ANONYMOUS_TRANSACTIONS toLOCAL or_`uuid`_, the replica must havegtid_mode=ON set, and this cannot be changed afterwards. This option is for use with a source that has binary log file position based replication, so SOURCE_AUTO_POSITION=1 cannot be set for the channel. Both the replication SQL thread and the replication I/O (receiver) thread must be stopped before setting this option.
  Important
  A replica set up withASSIGN_GTIDS_TO_ANONYMOUS_TRANSACTIONS on any channel cannot be promoted to replace the replication source server in the event that a failover is required, and a backup taken from the replica cannot be used to restore the replication source server. The same restriction applies to replacing or restoring other replicas that useASSIGN_GTIDS_TO_ANONYMOUS_TRANSACTIONS on any channel.
  For further restrictions and information, seeSection 19.1.3.6, “Replication From a Source Without GTIDs to a Replica With GTIDs”.
• GET_SOURCE_PUBLIC_KEY = {0|1}
  Enables RSA key pair-based password exchange by requesting the public key from the source. The option is disabled by default.
  This option applies to replicas that authenticate with thecaching_sha2_password authentication plugin. For connections by accounts that authenticate using this plugin, the source does not send the public key unless requested, so it must be requested or specified in the client. If SOURCE_PUBLIC_KEY_PATH is given and specifies a valid public key file, it takes precedence over GET_SOURCE_PUBLIC_KEY. If you are using a replication user account that authenticates with the caching_sha2_password plugin (which is the default from MySQL 8.0), and you are not using a secure connection, you must specify either this option or the SOURCE_PUBLIC_KEY_PATH option to provide the RSA public key to the replica.
• GTID_ONLY = {0|1}
  Stops the replication channel persisting file names and file positions in the replication metadata repositories.GTID_ONLY is available as of MySQL 8.0.27. The GTID_ONLY option is disabled by default for asynchronous replication channels, but it is enabled by default for Group Replication channels, and it cannot be disabled for them.
  For replication channels with this setting, in-memory file positions are still tracked, and file positions can still be observed for debugging purposes in error messages and through interfaces such as SHOW REPLICA STATUS statements (where they are shown as being invalid if they are out of date). However, the writes and reads required to persist and check the file positions are avoided in situations where GTID-based replication does not actually require them, including the transaction queuing and application process.
  This option can be used only if both the replication SQL (applier) thread and replication I/O (receiver) thread are stopped. To set GTID_ONLY = 1 for a replication channel, GTIDs must be in use on the server (gtid_mode = ON), and row-based binary logging must be in use on the source (statement-based replication is not supported). The optionsREQUIRE_ROW_FORMAT = 1 andSOURCE_AUTO_POSITION = 1 must be set for the replication channel.
  When GTID_ONLY = 1 is set, the replica usesreplica_parallel_workers=1 if that system variable is set to zero for the server, so it is always technically a multi-threaded applier. This is because a multi-threaded applier uses saved positions rather than the replication metadata repositories to locate the start of a transaction that it needs to reapply.
  If you disable GTID_ONLY after setting it, the existing relay logs are deleted and the existing known binary log file positions are persisted, even if they are stale. The file positions for the binary log and relay log in the replication metadata repositories might be invalid, and a warning is returned if this is the case. Provided that SOURCE_AUTO_POSITION is still enabled, GTID auto-positioning is used to provide the correct positioning.
  If you also disable SOURCE_AUTO_POSITION, the file positions for the binary log and relay log in the replication metadata repositories are used for positioning if they are valid. If they are marked as invalid, you must provide a valid binary log file name and position (SOURCE_LOG_FILE andSOURCE_LOG_POS). If you also provide a relay log file name and position (RELAY_LOG_FILE andRELAY_LOG_POS), the relay logs are preserved and the applier position is set to the stated position. GTID auto-skip ensures that any transactions already applied are skipped even if the eventual applier position is not correct.
• IGNORE_SERVER_IDS = (server_id_list)
  Makes the replica ignore events originating from the specified servers. The option takes a comma-separated list of 0 or more server IDs. Log rotation and deletion events from the servers are not ignored, and are recorded in the relay log.
  In circular replication, the originating server normally acts as the terminator of its own events, so that they are not applied more than once. Thus, this option is useful in circular replication when one of the servers in the circle is removed. Suppose that you have a circular replication setup with 4 servers, having server IDs 1, 2, 3, and 4, and server 3 fails. When bridging the gap by starting replication from server 2 to server 4, you can includeIGNORE_SERVER_IDS = (3) in theCHANGE REPLICATION SOURCE TO statement that you issue on server 4 to tell it to use server 2 as its source instead of server 3. Doing so causes it to ignore and not to propagate any statements that originated with the server that is no longer in use.
  If IGNORE_SERVER_IDS contains the server's own ID and the server was started with the--replicate-same-server-id option enabled, an error results.
  Note
  When global transaction identifiers (GTIDs) are used for replication, transactions that have already been applied are automatically ignored, so theIGNORE_SERVER_IDS function is not required and is deprecated. Ifgtid_mode=ON is set for the server, a deprecation warning is issued if you include the IGNORE_SERVER_IDS option in aCHANGE REPLICATION SOURCE TO statement.
  The source metadata repository and the output ofSHOW REPLICA STATUS provide the list of servers that are currently ignored. For more information, seeSection 19.2.4.2, “Replication Metadata Repositories”, andSection 15.7.7.35, “SHOW REPLICA STATUS Statement”.
  If a CHANGE REPLICATION SOURCE TO statement is issued without anyIGNORE_SERVER_IDS option, any existing list is preserved. To clear the list of ignored servers, it is necessary to use the option with an empty list:

CHANGE REPLICATION SOURCE TO IGNORE_SERVER_IDS = ();  

RESET REPLICA ALL clears IGNORE_SERVER_IDS.
Note
A deprecation warning is issued if SET GTID_MODE=ON is issued when any channel has existing server IDs set withIGNORE_SERVER_IDS. Before starting GTID-based replication, check for and clear all ignored server ID lists on the servers involved. TheSHOW REPLICA STATUS statement displays the list of ignored IDs, if there is one. If you do receive the deprecation warning, you can still clear a list aftergtid_mode=ON is set by issuing a CHANGE REPLICATION SOURCE TO statement containing theIGNORE_SERVER_IDS option with an empty list.

• NETWORK_NAMESPACE = 'namespace'
  The network namespace to use for TCP/IP connections to the replication source server or, if the MySQL communication stack is in use, for Group Replication’s group communication connections. The maximum length of the string value is 64 characters. If this option is omitted, connections from the replica use the default (global) namespace. On platforms that do not implement network namespace support, failure occurs when the replica attempts to connect to the source. For information about network namespaces, see Section 7.1.14, “Network Namespace Support”.NETWORK_NAMESPACE is available as of MySQL 8.0.22.
• PRIVILEGE_CHECKS_USER = {NULL | 'account'}
  Names a user account that supplies a security context for the specified channel. NULL, which is the default, means no security context is used.PRIVILEGE_CHECKS_USER is available as of MySQL 8.0.18.
  The user name and host name for the user account must follow the syntax described in Section 8.2.4, “Specifying Account Names”, and the user must not be an anonymous user (with a blank user name) or the CURRENT_USER. The account must have theREPLICATION_APPLIER privilege, plus the required privileges to execute the transactions replicated on the channel. For details of the privileges required by the account, seeSection 19.3.3, “Replication Privilege Checks”. When you restart the replication channel, the privilege checks are applied from that point on. If you do not specify a channel and no other channels exist, the statement is applied to the default channel.
  The use of row-based binary logging is strongly recommended when PRIVILEGE_CHECKS_USER is set, and you can set REQUIRE_ROW_FORMAT to enforce this. For example, to start privilege checks on the channelchannel_1 on a running replica, issue the following statements:

STOP REPLICA FOR CHANNEL 'channel_1';  
CHANGE REPLICATION SOURCE TO  
    PRIVILEGE_CHECKS_USER = 'user'@'host',  
    REQUIRE_ROW_FORMAT = 1,  
    FOR CHANNEL 'channel_1';  
START REPLICA FOR CHANNEL 'channel_1';  

• RELAY_LOG_FILE = 'relay_log_file' ,RELAY_LOG_POS = 'relay_log_pos'
  The relay log file name, and the location in that file, at which the replication SQL thread begins reading from the replica's relay log the next time the thread starts.RELAY_LOG_FILE can use either an absolute or relative path, and uses the same base name asSOURCE_LOG_FILE. The maximum length of the string value is 511 characters.
  A CHANGE REPLICATION SOURCE TO statement using RELAY_LOG_FILE,RELAY_LOG_POS, or both options can be executed on a running replica when the replication SQL (applier) thread is stopped. Relay logs are preserved if at least one of the replication applier thread and the replication I/O (receiver) thread is running. If both threads are stopped, all relay log files are deleted unless at least one of RELAY_LOG_FILE orRELAY_LOG_POS is specified. For the Group Replication applier channel (group_replication_applier), which only has an applier thread and no receiver thread, this is the case if the applier thread is stopped, but with that channel you cannot use the RELAY_LOG_FILE andRELAY_LOG_POS options.
• REQUIRE_ROW_FORMAT = {0|1}
  Permits only row-based replication events to be processed by the replication channel. This option prevents the replication applier from taking actions such as creating temporary tables and executing LOAD DATA INFILE requests, which increases the security of the channel. The REQUIRE_ROW_FORMAT option is disabled by default for asynchronous replication channels, but it is enabled by default for Group Replication channels, and it cannot be disabled for them. For more information, seeSection 19.3.3, “Replication Privilege Checks”.REQUIRE_ROW_FORMAT is available as of MySQL 8.0.19.
• REQUIRE_TABLE_PRIMARY_KEY_CHECK = {STREAM | ON | OFF | GENERATE}
  Available as of MySQL 8.0.20, this option lets a replica set its own policy for primary key checks, as follows:
  • ON: The replica setssql_require_primary_key = ON; any replicated CREATE TABLE or ALTER TABLE statement must result in a table that contains a primary key.
  • OFF: The replica setssql_require_primary_key = OFF; no replicated CREATE TABLE orALTER TABLE statement is checked for the presence of a primary key.
  • STREAM: The replica uses whatever value of sql_require_primary_key is replicated from the source for each transaction. This is the default value, and the default behavior.
  • GENERATE: Added in MySQL 8.0.32, this causes the replica to generate an invisible primary key for any InnoDB table that, as replicated, lacks a primary key. SeeSection 15.1.20.11, “Generated Invisible Primary Keys”, for more information.
    GENERATE is not compatible with Group Replication; you can use ON,OFF, or STREAM.
    A divergence based on the presence of a generated invisible primary key solely on a source or replica table is supported by MySQL Replication as long as the source supports GIPKs (MySQL 8.0.30 and later) and the replica uses MySQL version 8.0.32 or later. If you use GIPKs on a replica and replicate from a source using MySQL 8.0.29 or earlier, you should be aware that in this case such divergences in schema, other than the extra GIPK on the replica, are not supported and may result in replication errors.
    For multisource replication, settingREQUIRE_TABLE_PRIMARY_KEY_CHECK toON or OFF lets the replica normalize behavior across replication channels for different sources, and to keep a consistent setting forsql_require_primary_key. Using ON safeguards against the accidental loss of primary keys when multiple sources update the same set of tables. Using OFF lets sources that can manipulate primary keys to work alongside sources that cannot.
    In the case of multiple replicas, whenREQUIRE_TABLE_PRIMARY_KEY_CHECK is set toGENERATE, the generated invisible primary key added by a given replica is independent of any such key added on any other replica. This means that, if generated invisible primary keys are in use, the values in the generated primary key columns on different replicas are not guaranteed to be the same. This may be an issue when failing over to such a replica.
    When PRIVILEGE_CHECKS_USER isNULL (the default), the user account does not need administration level privileges to set restricted session variables. Setting this option to a value other thanNULL means that, whenREQUIRE_TABLE_PRIMARY_KEY_CHECK isON, OFF, orGENERATE, the user account does not require session administration level privileges to set restricted session variables such assql_require_primary_key, avoiding the need to grant the account such privileges. For more information, seeSection 19.3.3, “Replication Privilege Checks”.
• SOURCE_AUTO_POSITION = {0|1}
  Makes the replica attempt to connect to the source using the auto-positioning feature of GTID-based replication, rather than a binary log file based position. This option is used to start a replica using GTID-based replication. The default is 0, meaning that GTID auto-positioning and GTID-based replication are not used. This option can be used withCHANGE REPLICATION SOURCE TO only if both the replication SQL (applier) thread and replication I/O (receiver) thread are stopped.
  Both the replica and the source must have GTIDs enabled (GTID_MODE=ON,ON_PERMISSIVE, orOFF_PERMISSIVE on the replica, andGTID_MODE=ON on the source). SOURCE_LOG_FILE,SOURCE_LOG_POS,RELAY_LOG_FILE, andRELAY_LOG_POS cannot be specified together with SOURCE_AUTO_POSITION = 1. If multi-source replication is enabled on the replica, you need to set the SOURCE_AUTO_POSITION = 1 option for each applicable replication channel.
  With SOURCE_AUTO_POSITION = 1 set, in the initial connection handshake, the replica sends a GTID set containing the transactions that it has already received, committed, or both. The source responds by sending all transactions recorded in its binary log whose GTID is not included in the GTID set sent by the replica. This exchange ensures that the source only sends the transactions with a GTID that the replica has not already recorded or committed. If the replica receives transactions from more than one source, as in the case of a diamond topology, the auto-skip function ensures that the transactions are not applied twice. For details of how the GTID set sent by the replica is computed, seeSection 19.1.3.3, “GTID Auto-Positioning”.
  If any of the transactions that should be sent by the source have been purged from the source's binary log, or added to the set of GTIDs in thegtid_purged system variable by another method, the source sends the errorER_SOURCE_HAS_PURGED_REQUIRED_GTIDS to the replica, and replication does not start. The GTIDs of the missing purged transactions are identified and listed in the source's error log in the warning messageER_FOUND_MISSING_GTIDS. Also, if during the exchange of transactions it is found that the replica has recorded or committed transactions with the source's UUID in the GTID, but the source itself has not committed them, the source sends the errorER_REPLICA_HAS_MORE_GTIDS_THAN_SOURCE to the replica and replication does not start. For information on how to handle these situations, seeSection 19.1.3.3, “GTID Auto-Positioning”.
  You can see whether replication is running with GTID auto-positioning enabled by checking the Performance Schemareplication_connection_status table or the output ofSHOW REPLICA STATUS. Disabling theSOURCE_AUTO_POSITION option again makes the replica revert to file-based replication.
• SOURCE_BIND = 'interface_name'
  Determines which of the replica's network interfaces is chosen for connecting to the source, for use on replicas that have multiple network interfaces. Specify the IP address of the network interface. The maximum length of the string value is 255 characters.
  The IP address configured with this option, if any, can be seen in the Source_Bind column of the output fromSHOW REPLICA STATUS. In the source metadata repository table mysql.slave_master_info, the value can be seen as the Source_bind column. The ability to bind a replica to a specific network interface is also supported by NDB Cluster.
• SOURCE_COMPRESSION_ALGORITHMS = 'algorithm[,algorithm][,algorithm]'
  Specifies one, two, or three of the permitted compression algorithms for connections to the replication source server, separated by commas. The maximum length of the string value is 99 characters. The default value isuncompressed.
  The available algorithms are zlib,zstd, anduncompressed, the same as for theprotocol_compression_algorithms system variable. The algorithms can be specified in any order, but it is not an order of preference - the algorithm negotiation process attempts to use zlib, then zstd, thenuncompressed, if they are specified.SOURCE_COMPRESSION_ALGORITHMS is available as of MySQL 8.0.18.
  The value ofSOURCE_COMPRESSION_ALGORITHMS applies only if thereplica_compressed_protocol orslave_compressed_protocol system variable is disabled. Ifreplica_compressed_protocol orslave_compressed_protocol is enabled, it takes precedence overSOURCE_COMPRESSION_ALGORITHMS and connections to the source use zlib compression if both source and replica support that algorithm. For more information, seeSection 6.2.8, “Connection Compression Control”.
  Binary log transaction compression (available as of MySQL 8.0.20), which is activated by thebinlog_transaction_compression system variable, can also be used to save bandwidth. If you do this in combination with connection compression, connection compression has less opportunity to act on the data, but can still compress headers and those events and transaction payloads that are uncompressed. For more information on binary log transaction compression, seeSection 7.4.4.5, “Binary Log Transaction Compression”.
• SOURCE_CONNECT_RETRY =interval
  Specifies the interval in seconds between the reconnection attempts that the replica makes after the connection to the source times out. The default interval is 60.
  The number of attempts is limited by theSOURCE_RETRY_COUNT option. If both the default settings are used, the replica waits 60 seconds between reconnection attempts (SOURCE_CONNECT_RETRY=60), and keeps attempting to reconnect at this rate for 60 days (SOURCE_RETRY_COUNT=86400). These values are recorded in the source metadata repository and shown in thereplication_connection_configuration Performance Schema table.
• SOURCE_CONNECTION_AUTO_FAILOVER = {0|1}
  Activates the asynchronous connection failover mechanism for a replication channel if one or more alternative replication source servers are available (so when there are multiple MySQL servers or groups of servers that share the replicated data). SOURCE_CONNECTION_AUTO_FAILOVER is available as of MySQL 8.0.22. The default is 0, meaning that the mechanism is not activated. For full information and instructions to set up this feature, seeSection 19.4.9.2, “Asynchronous Connection Failover for Replicas”.
  The asynchronous connection failover mechanism takes over after the reconnection attempts controlled bySOURCE_CONNECT_RETRY andSOURCE_RETRY_COUNT are exhausted. It reconnects the replica to an alternative source chosen from a specified source list, which you manage using theasynchronous_connection_failover_add_source andasynchronous_connection_failover_delete_source functions. To add and remove managed groups of servers, use theasynchronous_connection_failover_add_managed andasynchronous_connection_failover_delete_managed functions instead. For more information, seeSection 19.4.9, “Switching Sources and Replicas with Asynchronous Connection Failover”.
  Important
  1. You can only setSOURCE_CONNECTION_AUTO_FAILOVER = 1 when GTID auto-positioning is in use (SOURCE_AUTO_POSITION = 1).
  2. When you set SOURCE_CONNECTION_AUTO_FAILOVER = 1, setSOURCE_RETRY_COUNT andSOURCE_CONNECT_RETRY to minimal numbers that just allow a few retry attempts with the same source, in case the connection failure is caused by a transient network outage. Otherwise the asynchronous connection failover mechanism cannot be activated promptly. Suitable values areSOURCE_RETRY_COUNT=3 andSOURCE_CONNECT_RETRY=10, which make the replica retry the connection 3 times with 10-second intervals between.
  3. When you set SOURCE_CONNECTION_AUTO_FAILOVER = 1, the replication metadata repositories must contain the credentials for a replication user account that can be used to connect to all the servers on the source list for the replication channel. The account must also have SELECT permissions on the Performance Schema tables. These credentials can be set using theCHANGE REPLICATION SOURCE TO statement with theSOURCE_USER andSOURCE_PASSWORD options. For more information, seeSection 19.4.9, “Switching Sources and Replicas with Asynchronous Connection Failover”.
  4. From MySQL 8.0.27, when you setSOURCE_CONNECTION_AUTO_FAILOVER = 1, asynchronous connection failover for replicas is automatically activated if this replication channel is on a Group Replication primary in a group in single-primary mode. With this function active, if the primary that is replicating goes offline or into an error state, the new primary starts replication on the same channel when it is elected. If you want to use the function, this replication channel must also be set up on all the secondary servers in the replication group, and on any new joining members. (If the servers are provisioned using MySQL’s clone functionality, this all happens automatically.) If you do not want to use the function, disable it by using thegroup_replication_disable_member_action() function to disable the Group Replication member actionmysql_start_failover_channels_if_primary, which is enabled by default. For more information, seeSection 19.4.9.2, “Asynchronous Connection Failover for Replicas”.
• SOURCE_DELAY =interval
  Specifies how many seconds behind the source the replica must lag. An event received from the source is not executed until at least interval seconds later than its execution on the source.interval must be a nonnegative integer in the range from 0 to 231−1. The default is 0. For more information, seeSection 19.4.11, “Delayed Replication”.
  A CHANGE REPLICATION SOURCE TO statement using the SOURCE_DELAY option can be executed on a running replica when the replication SQL thread is stopped.
• SOURCE_HEARTBEAT_PERIOD =interval
  Controls the heartbeat interval, which stops the connection timeout occurring in the absence of data if the connection is still good. A heartbeat signal is sent to the replica after that number of seconds, and the waiting period is reset whenever the source's binary log is updated with an event. Heartbeats are therefore sent by the source only if there are no unsent events in the binary log file for a period longer than this.
  The heartbeat interval interval is a decimal value having the range 0 to 4294967 seconds and a resolution in milliseconds; the smallest nonzero value is 0.001. Setting interval to 0 disables heartbeats altogether. The heartbeat interval defaults to half the value of thereplica_net_timeout orslave_net_timeout system variable. It is recorded in the source metadata repository and shown in thereplication_connection_configuration Performance Schema table.
  The system variablereplica_net_timeout (from MySQL 8.0.26) orslave_net_timeout (before MySQL 8.0.26) specifies the number of seconds that the replica waits for either more data or a heartbeat signal from the source, before the replica considers the connection broken, aborts the read, and tries to reconnect. The default value is 60 seconds (one minute). Note that a change to the value or default setting ofreplica_net_timeout orslave_net_timeout does not automatically change the heartbeat interval, whether that has been set explicitly or is using a previously calculated default. A warning is issued if you set the global value ofreplica_net_timeout orslave_net_timeout to a value less than that of the current heartbeat interval. Ifreplica_net_timeout orslave_net_timeout is changed, you must also issue CHANGE REPLICATION SOURCE TO to adjust the heartbeat interval to an appropriate value so that the heartbeat signal occurs before the connection timeout. If you do not do this, the heartbeat signal has no effect, and if no data is received from the source, the replica can make repeated reconnection attempts, creating zombie dump threads.
• SOURCE_HOST = 'host_name'
  The host name or IP address of the replication source server. The replica uses this to connect to the source. The maximum length of the string value is 255 characters.
  If you specify SOURCE_HOST orSOURCE_PORT, the replica assumes that the source server is different from before (even if the option value is the same as its current value.) In this case, the old values for the source's binary log file name and position are considered no longer applicable, so if you do not specify SOURCE_LOG_FILE andSOURCE_LOG_POS in the statement,SOURCE_LOG_FILE='' andSOURCE_LOG_POS=4 are silently appended to it.
  Setting SOURCE_HOST='' (that is, setting its value explicitly to an empty string) is_not_ the same as not settingSOURCE_HOST at all. Trying to setSOURCE_HOST to an empty string fails with an error.
• SOURCE_LOG_FILE = 'source_log_name',SOURCE_LOG_POS =source_log_pos
  The binary log file name, and the location in that file, at which the replication I/O (receiver) thread begins reading from the source's binary log the next time the thread starts. Specify these options if you are using binary log file position based replication.
  SOURCE_LOG_FILE must include the numeric suffix of a specific binary log file that is available on the source server, for example,SOURCE_LOG_FILE='binlog.000145'. The maximum length of the string value is 511 characters.
  SOURCE_LOG_POS is the numeric position for the replica to start reading in that file.SOURCE_LOG_POS=4 represents the start of the events in a binary log file.
  If you specify either of SOURCE_LOG_FILE or SOURCE_LOG_POS, you cannot specifySOURCE_AUTO_POSITION = 1, which is for GTID-based replication.
  If neither of SOURCE_LOG_FILE orSOURCE_LOG_POS is specified, the replica uses the last coordinates of the replication SQL thread before CHANGE REPLICATION SOURCE TO was issued. This ensures that there is no discontinuity in replication, even if the replication SQL (applier) thread was late compared to the replication I/O (receiver) thread.
• SOURCE_PASSWORD = 'password'
  The password for the replication user account to use for connecting to the replication source server. The maximum length of the string value is 32 characters. If you specifySOURCE_PASSWORD,SOURCE_USER is also required.
  The password used for a replication user account in aCHANGE REPLICATION SOURCE TO statement is limited to 32 characters in length. Trying to use a password of more than 32 characters causes CHANGE REPLICATION SOURCE TO to fail.
  The password is masked in MySQL Server’s logs, Performance Schema tables, and SHOW PROCESSLIST statements.
• SOURCE_PORT =port_num
  The TCP/IP port number that the replica uses to connect to the replication source server.
  Note
  Replication cannot use Unix socket files. You must be able to connect to the replication source server using TCP/IP.
  If you specify SOURCE_HOST orSOURCE_PORT, the replica assumes that the source server is different from before (even if the option value is the same as its current value.) In this case, the old values for the source's binary log file name and position are considered no longer applicable, so if you do not specify SOURCE_LOG_FILE andSOURCE_LOG_POS in the statement,SOURCE_LOG_FILE='' andSOURCE_LOG_POS=4 are silently appended to it.
• SOURCE_PUBLIC_KEY_PATH = 'key_file_name'
  Enables RSA key pair-based password exchange by providing the path name to a file containing a replica-side copy of the public key required by the source. The file must be in PEM format. The maximum length of the string value is 511 characters.
  This option applies to replicas that authenticate with thesha256_password orcaching_sha2_password authentication plugin. (For sha256_password,SOURCE_PUBLIC_KEY_PATH can be used only if MySQL was built using OpenSSL.) If you are using a replication user account that authenticates with thecaching_sha2_password plugin (which is the default from MySQL 8.0), and you are not using a secure connection, you must specify either this option or theGET_SOURCE_PUBLIC_KEY=1 option to provide the RSA public key to the replica.
• SOURCE_RETRY_COUNT =count
  Sets the maximum number of reconnection attempts that the replica makes after the connection to the source times out, as determined by thereplica_net_timeout orslave_net_timeout system variable. If the replica does need to reconnect, the first retry occurs immediately after the timeout. The default is 86400 attempts.
  The interval between the attempts is specified by theSOURCE_CONNECT_RETRY option. If both the default settings are used, the replica waits 60 seconds between reconnection attempts (SOURCE_CONNECT_RETRY=60), and keeps attempting to reconnect at this rate for 60 days (SOURCE_RETRY_COUNT=86400). A setting of 0 for SOURCE_RETRY_COUNT means that there is no limit on the number of reconnection attempts, so the replica keeps trying to reconnect indefinitely.
  The values for SOURCE_CONNECT_RETRY andSOURCE_RETRY_COUNT are recorded in the source metadata repository and shown in thereplication_connection_configuration Performance Schema table.SOURCE_RETRY_COUNT supersedes the--master-retry-count server startup option.
• SOURCE_SSL = {0|1}
  Specify whether the replica encrypts the replication connection. The default is 0, meaning that the replica does not encrypt the replication connection. If you setSOURCE_SSL=1, you can configure the encryption using theSOURCE_SSL_ _`xxx`_ andSOURCE_TLS_ _`xxx`_ options.
  Setting SOURCE_SSL=1 for a replication connection and then setting no furtherSOURCE_SSL_ _`xxx`_ options corresponds to setting--ssl-mode=REQUIRED for the client, as described inCommand Options for Encrypted Connections. With SOURCE_SSL=1, the connection attempt only succeeds if an encrypted connection can be established. A replication connection does not fall back to an unencrypted connection, so there is no setting corresponding to the --ssl-mode=PREFERRED setting for replication. If SOURCE_SSL=0 is set, this corresponds to --ssl-mode=DISABLED.
  Important
  To help prevent sophisticated man-in-the-middle attacks, it is important for the replica to verify the server’s identity. You can specify additionalSOURCE_SSL_ _`xxx`_ options to correspond to the settings--ssl-mode=VERIFY_CA and--ssl-mode=VERIFY_IDENTITY, which are a better choice than the default setting to help prevent this type of attack. With these settings, the replica checks that the server’s certificate is valid, and checks that the host name the replica is using matches the identity in the server’s certificate. To implement one of these levels of verification, you must first ensure that the CA certificate for the server is reliably available to the replica, otherwise availability issues will result. For this reason, they are not the default setting.
• SOURCE_SSL_ xxx,SOURCE_TLS_ xxx
  Specify how the replica uses encryption and ciphers to secure the replication connection. These options can be changed even on replicas that are compiled without SSL support. They are saved to the source metadata repository, but are ignored if the replica does not have SSL support enabled. The maximum length of the value for the string-valuedSOURCE_SSL_ _`xxx`_ andSOURCE_TLS_ _`xxx`_ options is 511 characters, with the exception ofSOURCE_TLS_CIPHERSUITES, for which it is 4000 characters.
  TheSOURCE_SSL_ _`xxx`_ andSOURCE_TLS_ _`xxx`_ options perform the same functions as the--ssl-_`xxx`_ and--tls-_`xxx`_ client options described inCommand Options for Encrypted Connections. The correspondence between the two sets of options, and the use of theSOURCE_SSL_ _`xxx`_ andSOURCE_TLS_ _`xxx`_ options to set up a secure connection, is explained inSection 19.3.1, “Setting Up Replication to Use Encrypted Connections”.
• SOURCE_USER = 'user_name'
  The user name for the replication user account to use for connecting to the replication source server. The maximum length of the string value is 96 characters.
  For Group Replication, this account must exist on every member of the replication group. It is used for distributed recovery if the XCom communication stack is in use for the group, and also used for group communication connections if the MySQL communication stack is in use for the group. With the MySQL communication stack, the account must have theGROUP_REPLICATION_STREAM permission.
  It is possible to set an empty user name by specifyingSOURCE_USER='', but the replication channel cannot be started with an empty user name. In releases before MySQL 8.0.21, only set an emptySOURCE_USER user name if you need to clear previously used credentials from the replication metadata repositories for security purposes. Do not use the channel afterwards, due to a bug in these releases that can substitute a default user name if an empty user name is read from the repositories (for example, during an automatic restart of a Group Replication channel). From MySQL 8.0.21, it is valid to set an empty SOURCE_USER user name and use the channel afterwards if you always provide user credentials using theSTART REPLICA statement or START GROUP_REPLICATION statement that starts the replication channel. This approach means that the replication channel always needs operator intervention to restart, but the user credentials are not recorded in the replication metadata repositories.
  Important
  To connect to the source using a replication user account that authenticates with thecaching_sha2_password plugin, you must either set up a secure connection as described inSection 19.3.1, “Setting Up Replication to Use Encrypted Connections”, or enable the unencrypted connection to support password exchange using an RSA key pair. Thecaching_sha2_password authentication plugin is the default for new users created from MySQL 8.0 (seeSection 8.4.1.2, “Caching SHA-2 Pluggable Authentication”). If the user account that you create or use for replication uses this authentication plugin, and you are not using a secure connection, you must enable RSA key pair-based password exchange for a successful connection. You can do this using either theSOURCE_PUBLIC_KEY_PATH option or theGET_SOURCE_PUBLIC_KEY=1 option for this statement.
• SOURCE_ZSTD_COMPRESSION_LEVEL =level
  The compression level to use for connections to the replication source server that use thezstd compression algorithm.SOURCE_ZSTD_COMPRESSION_LEVEL is available as of MySQL 8.0.18. The permitted levels are from 1 to 22, with larger values indicating increasing levels of compression. The default level is 3.
  The compression level setting has no effect on connections that do not use zstd compression. For more information, seeSection 6.2.8, “Connection Compression Control”.

Examples

CHANGE REPLICATION SOURCE TO is useful for setting up a replica when you have the snapshot of the source and have recorded the source's binary log coordinates corresponding to the time of the snapshot. After loading the snapshot into the replica to synchronize it with the source, you can run CHANGE REPLICATION SOURCE TO SOURCE_LOG_FILE='_`logname`_', SOURCE_LOG_POS=_`logpos`_ on the replica to specify the coordinates at which the replica should begin reading the source's binary log. The following example changes the source server the replica uses and establishes the source's binary log coordinates from which the replica begins reading:

CHANGE REPLICATION SOURCE TO
  SOURCE_HOST='source2.example.com',
  SOURCE_USER='replication',
  SOURCE_PASSWORD='password',
  SOURCE_PORT=3306,
  SOURCE_LOG_FILE='source2-bin.001',
  SOURCE_LOG_POS=4,
  SOURCE_CONNECT_RETRY=10;

For the procedure to switch an existing replica to a new source during failover, seeSection 19.4.8, “Switching Sources During Failover”.

When GTIDs are in use on the source and the replica, specify GTID auto-positioning instead of giving the binary log file position, as in the following example. For full instructions to configure and start GTID-based replication on new or stopped servers, online servers, or additional replicas, seeSection 19.1.3, “Replication with Global Transaction Identifiers”.

CHANGE REPLICATION SOURCE TO
  SOURCE_HOST='source3.example.com',
  SOURCE_USER='replication',
  SOURCE_PASSWORD='password',
  SOURCE_PORT=3306,
  SOURCE_AUTO_POSITION = 1,
  FOR CHANNEL "source_3";

In this example, multi-source replication is in use, and theCHANGE REPLICATION SOURCE TO statement is applied to the replication channel"source_3" that connects the replica to the specified host. For guidance on setting up multi-source replication, see Section 19.1.5, “MySQL Multi-Source Replication”.

The next example shows how to make the replica apply transactions from relay log files that you want to repeat. To do this, the source need not be reachable. You can useCHANGE REPLICATION SOURCE TO to locate the relay log position where you want the replica to start reapplying transactions, and then start the SQL thread:

CHANGE REPLICATION SOURCE TO
  RELAY_LOG_FILE='replica-relay-bin.006',
  RELAY_LOG_POS=4025;
START REPLICA SQL_THREAD;

CHANGE REPLICATION SOURCE TO can also be used to skip over transactions in the binary log that are causing replication to stop. The appropriate method to do this depends on whether GTIDs are in use or not. For instructions to skip transactions usingCHANGE REPLICATION SOURCE TO or another method, seeSection 19.1.7.3, “Skipping Transactions”.