I've installed RAC One Node for the first time and currently playing around with it to get better to know. The underlying cluster has two nodes, let's call them NODE_A and NODE_B.
The cluster database itself is called DEVELOP, instance DEVELOP_1 is running on node NODE_A, instance DEVELOP_2 is supposed to run on NODE_B. The preferred node is
When testing database online relcation everything works as expected: instance DEVELOP_1 gets shutdown on node NODE_A - instance DEVELOP_2 gets started on NODE_B.
But when shutting down instance DEVELOP_1 on NODE_A with SQL*Plus I was astonished about the result:
- it was not instance DEVELOP_2 which got started on NODE_B - it was instance DEVELOP_1! (Due to this my Coud Control got changed - but that's not the issue of this discussion).
My questions now are:
- is the clusterware responsible for the behaviour of starting instance DEVELOP_1 on NODE_B when I shut down instance DEVELOP_1 down on NODE_A via SQL*Plus?
- and were is this behaviour documented? I went through the RAC One Node documentation but couldn't find anything helpful
Any help will be appreciated.
Yes, its expected behaviour in RAC one node, please check the document below.
"Oracle Real Application Clusters One Node Overview " (Page-1) (Online database relocation)