In cluster environments, if we connect using connect string or JDBC, we usually connect to a public address that leads us to some of the nodes in the cluster. Sometimes We need to go to a particular node, from applications (of course, you can do this by defining preferred affinity and services) or as in my case for administration issues, where You need to work on a particular node.
Since version 11, if I am not wrong, the connect string includes the parameter instance_name, an example we can see below for a cluster of three nodes, with this we can go directly to the node we need.
sap_1.domain.es = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = scan.public.com)(PORT = 5671)) ) (CONNECT_DATA = (SERVICE_NAME = SAP) (INSTANCE_NAME=SAP_1) ) ) sap_2.domain.es = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = scan.public.com)(PORT = 5671)) ) (CONNECT_DATA = (SERVICE_NAME = SAP) (INSTANCE_NAME=SAP_2) ) ) sap_3.domain.es = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = scan.public.com)(PORT = 5671)) ) (CONNECT_DATA = (SERVICE_NAME = SAP) (INSTANCE_NAME=SAP_3) ) )
HTH – Antonio NAVARRO
Advertisements