Scale Out Oracle RAC Database as a Service with Robin Platform
Oracle RAC Database as a Service – How to Scale
See how easy it is for anybody to stand up an entirely new Oracle RAC environment including the grid infrastructure installation the ASM configuration and finally create the RAC database tool.
You have seen how easy it is to deploy a fresh new Oracle RAC database environment. But what if we want to know how our workload might respond when adding a third node to the cluster? In other words, test the scalability of that particular workload when adding a third node.
So it’s really easy. We just click on “Scale Out” for the application and here we can define the number of nodes by which we want to extend this cluster. This is done simply by sliding across this bar but for this demonstration purpose, we need to add a single node. We can also explicitly
call out a hostname for the new node. We can go back and tweak some of the environment variables as input for this new operation but for this demo, we really don’t need to make any of these changes.
So let’s just close these out and just simply click on the “Scale Out” button to begin the process for extending our RAC cluster. Behind the scenes, Robin is making all the necessary calls to Oracle to affect the extension of the cluster – in very much the same way as you might through conventional means for any other installation ensuring that from an Oracle perspective things are all agreeable with the configuration. You can see the success of the operation in this window. We’ll close this window and now we are back on our application screen with the newly refreshed view to find that our third node has been added.
We can see the new IP addresses – the physical host on which the new container has been deployed. Let’s just jump back into the new container – rather do a similar verification to see that we have actually successfully reshaped RAC database environment with three nodes from two nodes. We now log into Oracle set our environment through SRB CTL. Let’s just do a status again of our Robin database so we can see that we’ve got our Robin three instance now, which has been added and it’s now running on our new V node.
In the new container in the Robin cluster, we can see the new vip is added and is up and running. The resources have been successfully configured across the new node and if we go back into SQL plus and log back into the database itself and do once again a query of gv$instance, we can see that we had the databases up and fully available across all three instances of the cluster. Okay, so we exit out of that. We’re back to the UI and so now what if we want to scale back in? So we need to shrink that cluster – testing is completed – so we need to shrink that cluster back to two nodes –
Watch the demo to understand how to scale back.