Removing a node from replication
In some cases, a replicated node cannot be recovered quickly enough; that is, before other servers’ queues fill up.
In these cases, the configuration of still-healthy servers should be modified to stop replication with unavailable application nodes, as described below.
After finishing any changes remember to verify the replication configuration. See Cleaning up replication configuration
Configuring replacement servers
Any settings that the removed server was responsible for must be configured to run on a surviving server, including:
Service IDs in use by managed system policies.
Proxy server settings on target systems.
Reconfigure email settings.
If required, update links, load balancers, or network settings to direct users to log in using the new front-end server
Server responsibilities:
Use the servicemove program to move the failed server’s workload to the new server.
Removing a database node
To remove an application node if administrators still log into the Bravura Security Fabric interface:
Log into any one of the working servers as a product administrator with the Maintain servers administrative privilege.
Click Manage the system > Maintenance > Database replication.
Select the node that is causing the problem and click Delete.
If at least one server is still available to replicate to, click Propagate and reload replication configuration on all servers (without resynchronizing nodes) to copy the changes to all remaining Bravura Security Fabric servers and restart the database.
To remove a shared schema application node, after decommissioning the node see Cleaning up replication configuration .
If the shared schema node application services ever come back up and that node can connect to the database, the node will re-register itself and will show up in shared schema with the node that replaced it.
To remove an application node from a hybrid schema topology without replacing it, all application nodes that replicate to it from the other database nodes must also be removed. In effect, the reverse of adding an application node in shared schema has to be applied.