Databases Reference
In-Depth Information
12. Select the All servers in the cluster radio button from the
PROD_Cluster
cluster.
Click on the Finish button.
13. Finally, click on the Activate Changes button.
How it works...
The GridLink data source has some advantages over the multi data source.
The multi data source uses the test connection feature to guarantee a reliable connection
to the database. The GridLink data source on the other hand uses the call back mechanism
of Oracle RAC Fast Application Notification (FAN). This means that the GridLink data
source actively responds to events and notifications coming from the database, such as the
fluctuation of RAC services.
Another improvement is the load balancing mechanism. The multi data source configured with
the
Load Balance
algorithm distributes the application requests to borrow a connection
from the data source members of the multi data source in a round-robin fashion. GridLink
improves the distribution load by receiving load balancing events from the database. These
events indicate the recommended connection distribution among the RAC nodes.
If FAN is disabled, GridLink uses the round-robin
Load Balance
algorithm.
There's more...
The GridLink data source can also be created using WLST.
Creating a GridLink data source using WLST
1. Log in as a
wls
user to shell and start WLST:
[wls@prod01]$ $WL_HOME/common/bin/wlst.sh
2. Connect to the Administration Server using
wlsadmin
as the user,
<pwd>
as the
password, and
t3://adminhost.domain.local:7001
as the server URL:
wls:/offline> connect("wlsadmin","<pwd>","t3://adminhost.domain.
local:7001")
3. Run the following WLST commands to create the GridLink data source:
edit()
startEdit()
cmo.createJDBCSystemResource('ds-GridLinkXA')
Search WWH ::
Custom Search