Selectively Skipping Replication of Binlog Events
Learn techniques to bypass specific replication events. This guide explains how to ignore individual transactions or errors to restore replication flow after a stoppage.
The terms master and slave have historically been used in replication, and MariaDB has begun the process of adding primary and replica synonyms. The old terms will continue to be used to maintain backward compatibility - see MDEV-18777 to follow progress on this effort.
Normally, all changes that are logged as events in the binary log are also replicated to all replicas (though still subject to filtering by replicate-do-db, replicate-ignore-db, and similar options). However, sometimes it may be desirable to have certain events logged in the binlog, but not replicated to all or to some replicas, where the distinction between events that should be replicated or not is under the control of the application making the changes.
This can be useful if an application does some replication external to the server outside of the built-in replication, or if it has data that for some reason should not be replicated.
This is possible with the following system variables.
Primary Session Variable: skip_replication
When the skip_replication variable is set to true, changes are logged into the binary log with the flag @@skip_replication set. Such events are not replicated by replicas that run with --replicate-events-marked-for-skip set differently from the default (which is REPLICATE).
Scope
Session only
Access Type
Dynamic
Data Type
bool
Default Value
OFF
The skip_replication option only has effect if binary logging is enabled and sql_log_bin is true.
Attempting to change @@skip_replication in the middle of a transaction fails. This is to avoid getting half of a transaction replicated while the other half is not replicated. Be sure to end any current transaction withCOMMIT/ROLLBACK before changing the variable.
Replica Option: --replicate-events-marked-for-skip
The replicate_events_marked_for_skip option tells the replica whether to replicate events that are marked with the @@skip_replication flag. Default is REPLICATE, to ensure that all changes are replicated to the replica. If set to FILTER_ON_SLAVE, events so marked will be skipped on the replica and not replicated. If set to FILTER_ON_MASTER, the filtering is done on the primary, saving
network bandwidth as the events are not received by the replica at all.
Scope
Global
Access Type
Dynamic
Data Type
enum: REPLICATE | FILTER_ON_SLAVE | FILTER_ON_MASTER
Default Value
REPLICATE
replicate_events_marked_for_skip is a dynamic variable (it can be changed without restarting the server). However, the replica threads must be stopped when it is changed, otherwise an error is thrown.
When events are filtered due to @@skip_replication, the filtering happens on the primary side; in other words, the event is never sent to the replica. If many events are filtered like this, a replica can sit a long time without receiving any events from the primary. This is not a problem in itself, but must be kept in mind when inquiring on the replica about events that are filtered. For example, START SLAVE UNTIL <some position> stops when the first event that is not filtered is encountered at the given position or beyond. If the event at the given position is filtered, then the replica thread only stops when the next non-filtered event is encountered. In effect, if an event is filtered, to the replica it appears that it was never written to the binlog on the primary.
Note that when events are filtered for a replica, the data in the database is different on the replica and on the primary. It is the responsibility of the application to replicate the data outside of the built-in replication or otherwise ensure consistency of operation. If this is not done, it is possible for replication to encounter, for example, UNIQUE constraint violations or other problems which will cause replication to stop and require manual intervention to fix.
The session variable @@skip_replication can be changed without requiring special privileges. This makes it possible for normal applications to control it without requiring SUPER privileges. But it must be kept in mind when using replicas with --replicate-events-marked-for-skip set different from REPLICATE, as it allows any connection to do changes that are not replicated.
skip_replication and sql_log_bin
@@sql_log_bin and @@skip_replication are somewhat related, as they can both be used to prevent a change on the primary from being replicated to the replica. The difference is that with @@skip_replication, changes are still written into the binlog, and replication of the events is only skipped on replicas that explicitly are configured to do so, with --replicate-events-marked-for-skip different from REPLICATE. With @@sql_log_bin, events are not logged into the binlog, and so are not replicated by any replica.
skip_replication and the Binlog
When events in the binlog are marked with the @@skip_replication flag, the flag will be preserved if the events are dumped by the mariadb-binlog program and re-applied against a server with the mariadb client program. Similarly, the BINLOG statement will preserve the flag from the event being replayed. And a replica which runs with--log-slave-updates and does not filter events (--replicate-events-marked-for-skip=REPLICATE) also preserves the flag in the events logged into the binlog on the replica.
See Also
Using SQL_SLAVE_SKIP_COUNTER - How to skip a number of events on the replica
This page is licensed: CC BY-SA / Gnu FDL
Last updated
Was this helpful?

