Controlling Redundancy

After the redundancy feature has been configured and is enabled, you can control its behavior using the redundancy control commands:

CAUTION: Please carefully read and adhere to the following regulations. Failure to do so may result in the improper functioning of the Hot Standby Redundancy system:

1. Redundancy control commands MUST ONLY be executed on one site at a time.
2. Ensure that all redundancy control command operations (such as start, stop, restart, etc.) are fully completed before performing any other actions.
3. Contact our technical support team if any unexpected errors occur.

If you want to import a license, the license file must be present on the Legacy Orchestrator appliance. See Copying the License File to Both Sites.

• redundancy control set-preferred-site – Updates the preferred site. Effective as soon as it is entered. Does not require a redundancy restart.
  • Three options are available: none, site-a, site-b.
• redundancy control start – Activates redundancy feature. If successful, data replication/resource monitoring will start.
• redundancy control stop – Stops the redundancy feature. Replication, monitoring and failover functions are all stopped.
  Important:
  • Stopping redundancy will stop data replication. Data will be stored to the data store on the active site only.
  • The redundancy feature can not stop properly if there is any terminal that is accessing the replication-partition folder (/home/skylight/so/mysql-ha/). Ensure that the replication-partition folder (/home/skylight/so/mysql-ha/) is not busy while stopping redundancy feature.
  • Redundancy feature can be stopped on both sites by running redundancy control stop command on a site only.
  • However, the redundancy feature must be stopped by manually rerunning redundancy control stop command again on the remaining site if a site is in shutdown or unreachable while stopping the redundancy feature process.
  • Do not force reboot/shutdown Legacy Orchestrator (containers or Docker host) while stopping redundancy feature process.
• redundancy control restart – Equivalent to entering the start and stop commands. If you change the redundancy configuration, you must enter the restart command.
• redundancy control switch – Switches the active site (from the site that is currently active to the site that is currently passive). If a manual switch is initiated, the Preferred-site option will be automatically set to none.
• redundancy control suspend – This command puts the redundancy features completely on hold:
• It disables failover (automatic and manual).
• It stops replication and monitoring immediately.
• Failover will not occur regardless of the conditions.

To re-enable replication, monitoring and automatic failover, you must enter the resume command.
The suspend command is typically used to temporarily put the redundancy feature on hold. For example, during a maintenance window, when you need to reboot an appliance or apply a patch to the operating system.

• redundancy control resume – Re-enables automatic failover after the suspend command was executed.
• redundancy control handle-split-brain – Handling data cannot synchronize between two sites issue.

All commands are run from the socli. See Accessing the climanager service (socli).