START REPLICA [thread_types] [until_option] [connection_options] [channel_option]

thread_types:
    [thread_type [, thread_type] ... ]

thread_type:
    IO_THREAD | SQL_THREAD

until_option:
    UNTIL {   {SQL_BEFORE_GTIDS | SQL_AFTER_GTIDS} = gtid_set
          |   MASTER_LOG_FILE = 'log_name', MASTER_LOG_POS = log_pos
          |   SOURCE_LOG_FILE = 'log_name', SOURCE_LOG_POS = log_pos
          |   RELAY_LOG_FILE = 'log_name', RELAY_LOG_POS = log_pos
          |   SQL_AFTER_MTS_GAPS  }

connection_options:
    [USER='user_name'] [PASSWORD='user_pass'] [DEFAULT_AUTH='plugin_name'] [PLUGIN_DIR='plugin_dir']


channel_option:
    FOR CHANNEL channel

gtid_set:
    uuid_set [, uuid_set] ...
    | ''

uuid_set:
    uuid:interval[:interval]...

uuid:
    hhhhhhhh-hhhh-hhhh-hhhh-hhhhhhhhhhhh

h:
    [0-9,A-F]

interval:
    n[-n]

    (n >= 1)

START REPLICA starts one or both of the replication threads. From MySQL 8.0.22, use START REPLICA in place of START SLAVE, which is deprecated from that release. In releases before MySQL 8.0.22, use START SLAVE.

START REPLICA with no thread_type options starts both of the replication threads. The replication I/O (receiver) thread reads events from the source server and stores them in the relay log. The replication SQL (applier) thread reads events from the relay log and executes them. START REPLICA requires the REPLICATION_SLAVE_ADMIN privilege (or the deprecated SUPER privilege).

If START REPLICA succeeds in starting the replication threads, it returns without any error. However, even in that case, it might be that the replication threads start and then later stop (for example, because they do not manage to connect to the source or read its binary log, or some other problem). START REPLICA does not warn you about this. You must check the replica's error log for error messages generated by the replication threads, or check that they are running satisfactorily with SHOW REPLICA STATUS.

START REPLICA causes an implicit commit of an ongoing transaction. See Section 13.3.3, “Statements That Cause an Implicit Commit”.

gtid_next must be set to AUTOMATIC before issuing this statement.

The optional FOR CHANNEL channel clause enables you to name which replication channel the statement applies to. Providing a FOR CHANNEL channel clause applies the START REPLICA statement to a specific replication channel. If no clause is named and no extra channels exist, the statement applies to the default channel. If a START REPLICA statement does not have a channel defined when using multiple channels, this statement starts the specified threads for all channels. This statement is disallowed for the group_replication_recovery channel. See Section 17.2.2, “Replication Channels” for more information.

You can add IO_THREAD and SQL_THREAD options to the statement to name which of the threads to start. Note that the Group Replication applier channel (group_replication_applier) has no I/O (receiver) thread, only an SQL (applier) thread. Specifying the IO_THREAD or SQL_THREAD options when you start this channel has no benefit.

START REPLICA supports pluggable user-password authentication with the USER, PASSWORD, DEFAULT_AUTH and PLUGIN_DIR options, as described in the following list:

You cannot use the SQL_THREAD option when specifying any of USER, PASSWORD, DEFAULT_AUTH, or PLUGIN_DIR, unless the IO_THREAD option is also provided.

For more information, see Section 6.2.17, “Pluggable Authentication”.

If an insecure connection is used with any these options, the server issues the warning Sending passwords in plain text without SSL/TLS is extremely insecure.

START REPLICA ... UNTIL supports two additional options for use with global transaction identifiers (GTIDs) (see Section 17.1.3, “Replication with Global Transaction Identifiers”). Each of these takes a set of one or more global transaction identifiers gtid_set as an argument (see GTID Sets, for more information).

When no thread_type is specified, START REPLICA UNTIL SQL_BEFORE_GTIDS causes the replication SQL thread to process transactions until it has reached the first transaction whose GTID is listed in the gtid_set. START REPLICA UNTIL SQL_AFTER_GTIDS causes the replication threads to process all transactions until the last transaction in the gtid_set has been processed by both threads. In other words, START REPLICA UNTIL SQL_BEFORE_GTIDS causes the replication SQL thread to process all transactions occurring before the first GTID in the gtid_set is reached, and START REPLICA UNTIL SQL_AFTER_GTIDS causes the replication threads to handle all transactions, including those whose GTIDs are found in gtid_set, until each has encountered a transaction whose GTID is not part of the set. SQL_BEFORE_GTIDS and SQL_AFTER_GTIDS each support the SQL_THREAD and IO_THREAD options, although using IO_THREAD with them currently has no effect.

For example, START REPLICA SQL_THREAD UNTIL SQL_BEFORE_GTIDS = 3E11FA47-71CA-11E1-9E33-C80AA9429562:11-56 causes the replication SQL thread to process all transactions originating from the source whose server_uuid is 3E11FA47-71CA-11E1-9E33-C80AA9429562 until it encounters the transaction having sequence number 11; it then stops without processing this transaction. In other words, all transactions up to and including the transaction with sequence number 10 are processed. Executing START REPLICA SQL_THREAD UNTIL SQL_AFTER_GTIDS = 3E11FA47-71CA-11E1-9E33-C80AA9429562:11-56, on the other hand, would cause the replication SQL thread to obtain all transactions just mentioned from the source, including all of the transactions having the sequence numbers 11 through 56, and then to stop without processing any additional transactions; that is, the transaction having sequence number 56 would be the last transaction fetched by the replication SQL thread.

When using a multithreaded replica (replica_parallel_workers or slave_parallel_workers > 0), there is a chance of gaps in the sequence of transactions that have been executed from the relay log in the following cases:

From MySQL 8.0.27, all replicas are multithreaded by default. When replica_preserve_commit_order=ON or slave_preserve_commit_order=ON is set for the replica, which is also the default setting from MySQL 8.0.27, gaps should not occur except in the specific situations listed in the description for replica_preserve_commit_order and slave_preserve_commit_order. If replica_preserve_commit_order=OFF or slave_preserve_commit_order=OFF is set for the replica, which is the default before MySQL 8.0.27, the commit order of transactions is not preserved, so the chance of gaps occurring is much larger.

Use the START REPLICA UNTIL SQL_AFTER_MTS_GAPS statement to cause a multithreaded replica's worker threads to only run until no more gaps are found in the relay log, and then to stop. This statement can take an SQL_THREAD option, but the effects of the statement remain unchanged. It has no effect on the replication I/O (receiver) thread (and cannot be used with the IO_THREAD option).

Before MySQL 8.0.26, issuing START REPLICA on a multithreaded replica with gaps in the sequence of transactions executed from the relay log generates a warning. In such a situation, the solution is to use START REPLICA UNTIL SQL_AFTER_MTS_GAPS, then issue RESET REPLICA to remove any remaining relay logs. 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, START REPLICA UNTIL SQL_AFTER_MTS_GAPS just stops the applier thread when it finds the first event to execute, and does not attempt to check for gaps in the sequence of transactions. See Section 17.5.1.34, “Replication and Transaction Inconsistencies” for more information.

To change a failed multithreaded replica to single-threaded mode, you can issue the following series of statements, in the order shown:

START SLAVE UNTIL SQL_AFTER_MTS_GAPS;
SET @@GLOBAL.slave_parallel_workers = 0;
START SLAVE SQL_THREAD;

Or from MySQL 8.0.26:
START REPLICA UNTIL SQL_AFTER_MTS_GAPS;
SET @@GLOBAL.replica_parallel_workers = 0;
START REPLICA SQL_THREAD;

START REPLICA sends an acknowledgment to the user after both the replication I/O (receiver) thread and the replication SQL (applier) thread have started. However, the replication receiver thread may not yet have connected. For this reason, a successful START REPLICA causes SHOW REPLICA STATUS to show Replica_SQL_Running=Yes, but this does not guarantee that Replica_IO_Running=Yes (because Replica_IO_Running=Yes only if the receiver thread is running and connected). For more information, see Section 13.7.7.35, “SHOW REPLICA STATUS Statement”, and Section 17.1.7.1, “Checking Replication Status”.

An UNTIL clause (until_option, in the preceding grammar) may be added to specify that the replica should start and run until the replication SQL thread reaches a given point in the source's binary log or in the replica's relay log. Use one of the following pairs of options to specify the position:

For compressed transaction payloads, the position must be based on the compressed Transaction_payload_event. When the SQL thread reaches the point specified, it stops. If the SQL_THREAD option is specified in the statement, it starts only the SQL thread. Otherwise, it starts both replication threads. If the SQL thread is running, the UNTIL clause is ignored and a warning is issued. You cannot use an UNTIL clause with the IO_THREAD option.

It is also possible with START REPLICA UNTIL to specify a stop point relative to a given GTID or set of GTIDs using one of the options SQL_BEFORE_GTIDS or SQL_AFTER_GTIDS, as explained previously in this section. When using one of these options, you can specify SQL_THREAD, IO_THREAD, both of these, or neither of them. If you specify only SQL_THREAD, then only the replication SQL thread (applier) is affected by the statement; if only IO_THREAD is used, then only the replication I/O (receiver) thread is affected. If both SQL_THREAD and IO_THREAD are used, or if neither of them is used, then both the SQL (applier) and I/O (receiver) threads are affected by the statement.

For an UNTIL clause, you must specify any one of the following:

Do not mix source and relay log options. Do not mix log file options with GTID options.

The UNTIL clause can still be used with the file name and file position options when the GTID_ONLY option was set on the CHANGE REPLICATION SOURCE TO statement to stop the replication channel from persisting file names and file positions in the replication metadata repositories. The file names and file positions are tracked in memory.

The UNTIL clause is not supported for multithreaded replicas except when also using SQL_AFTER_MTS_GAPS. If UNTIL is used on a multithreaded replica without SQL_AFTER_MTS_GAPS, the replica operates in single-threaded (sequential) mode for replication until the point specified by the UNTIL clause is reached.

Any UNTIL condition is reset by a subsequent STOP REPLICA statement, a START REPLICA statement that includes no UNTIL clause, or a server restart.

When specifying a log file and position, you can use the IO_THREAD option with START REPLICA ... UNTIL even though only the SQL thread is affected by this statement. The IO_THREAD option is ignored in such cases. The preceding restriction does not apply when using one of the GTID options (SQL_BEFORE_GTIDS and SQL_AFTER_GTIDS); the GTID options support both SQL_THREAD and IO_THREAD, as explained previously in this section.

The UNTIL clause can be useful for debugging replication, or to cause replication to proceed until just before the point where you want to avoid having the replica replicate an event. For example, if an unwise DROP TABLE statement was executed on the source, you can use UNTIL to tell the replica to execute up to that point but no farther. To find what the event is, use mysqlbinlog with the source's binary log or the replica's relay log, or by using a SHOW BINLOG EVENTS statement.

If you are using UNTIL to have the replica process replicated queries in sections, it is recommended that you start the replica with the --skip-slave-start option, or from MySQL 8.0.24, the skip_slave_start system variable, to prevent the SQL thread from running when the replica server starts. Remove the option or system variable setting after the procedure is complete, so that it is not forgotten in the event of an unexpected server restart.

The SHOW REPLICA STATUS statement includes output fields that display the current values of the UNTIL condition.