2. Home
  4. Docs
  6. ClusterControl
  8. User Guide (CLI)
  10. s9s-replication

s9s-replication

Manage database replication related functions.

Note

Only applicable for supported database clusters namely MySQL/MariaDB Replication and PostgreSQL Streaming Replication.

Usage

s9s replication {command} {options}

Command

Name, shorthand Description
−−failover Takes over the role of master from a failed master.
−−list List the replication links.
−−promote Promotes a slave to become a master.
−−stage Rebuilds or stages a replication slave.
−−start Starts a replication slave previously stopped using the --stop option.
−−stop Make the slave stop replicating. This option will create a job that does not stop the server but stops the replication on it.

Options

Name, shorthand Description
−−fail-stop-slave When this option is specified slave node will be stopped.
−−link-format=FORMATSTRING The format string controls the format of the printed information about the replication links. See Link Format.
−−master=NODE The replication master.
−−remote-cluster-id=ID Remote cluster ID for the c2c replication.
−−replication-master=NODE This is the same as the --master option.
−−slave=NODE The replication slave.
−−replication-slave=NODE This is the same as the --slave option.

Link Format

The format string controls the format of the printed information about the links. When this command-line option is used the specified information will be printed instead of the default columns. The format string uses the % character to mark variable fields and flag characters as they are specified in the standard printf() C library functions. The % specifiers are ended by field name letters to refer to various properties of the replication.

The %+12p format string for example has the +12 flag characters in it with the standard meaning: the field will be 12 characters wide and the + or - the sign will always be printed with the number. The properties of the links are encoded by letters. The in the %4p for example the letter p encodes the slave port field, so the port number of the slave node will be substituted. Standard \ the notation is also available, \n for example encodes a new-line character.

The following example prints out a customized format to show the replication link among nodes together with the position of master/slave in the replication log for PostgreSQL streaming replication, MySQL GTID and MariaDB GTID:

$ s9s replication \
        --list \
        --long \
        --link-format="%16h %4p <- %H %2P %o %O\n"
192.168.0.81 5432 <- 192.168.0.83 5432 0/71001BA8
192.168.0.82 5432 <- 192.168.0.83 5432 0/71001BA8
192.168.0.42 3306 <- 192.168.0.41 3306 dca7e205-90db-11ea-9143-5254008afee6:1-10 3157
192.168.0.43 3306 <- 192.168.0.41 3306 dca7e205-90db-11ea-9143-5254008afee6:1-10 3157
192.168.0.92 3306 <- 192.168.0.91 3306 0-54001-49 10285
192.168.0.93 3306 <- 192.168.0.91 3306 0-54001-49 10285

The s9s-tools support the following fields:

Field Description
c The cluster ID of the cluster where the slave node can be found.
C The master cluster-ID property of the slave host. This shows in which cluster the master of the represented replication link can be found.
d This format specifier denotes the “seconds behind the master” property of the slave.
h The hostname of the slave node in the link.
H The hostname of the master node in the link.
o The position of the slave in the replication log.
O The position of the master in the replication log.
p The port number of the slave.
P The port number of the master node.
s A short string representing the link status, e.g. “Online” when everything is ok.
m A slightly longer, human-readable string representing the state of the link. This is actually the slave_io_state property of the slave node.

Examples

Print all database replication links for clusters that fall under the replication category:

$ s9s replication --list

Promote a slave 192.168.0.164 to become a new master for a database cluster named “MySQL Rep 5.7 – Production”:

$ s9s replication \
        --promote \
        --cluster-name="MySQL Rep 5.7 - Production" \
        --slave="192.168.0.164:3306" \
        --wait

Rebuild the replication slave of 192.168.0.83 from the master, 192.168.0.76, and tag the job as “stage” for cluster ID 1:

$ s9s replication \
        --stage \
        --cluster-id=1 \
        --job-tags="stage" \
        --slave="192.168.0.83:3306" \
        --master="192.168.0.76:3306" \
        --wait

Stop the replication slave on node 192.168.0.80:

$ s9s replication \
        --stop \
        --cluster-id=1 \
        --slave="192.168.0.80:3306" \
        --wait

Start the replication slave on node 192.168.0.80:

$ s9s replication \
        --start \
        --cluster-id=1 \
        --slave="192.168.0.80:3306" \
        --wait
Was this article helpful to you? Yes No