I have a cluster, call it CLUSTERA, containing SERVER1 and SERVER2. Then I have some other servers which don't need names for the purposes of my question here. I am trying to build queries across these servers and I have set up the appropriate linked servers, which are working.
What I would like to do is address each database table by a four-part name so that I can execute the same query from any server:
[blah blah] FROM CLUSTERA.mydatabasename.schema.table INNER JOIN otherserver.thatdatabase.schema.table [et cetera]
However, when I execute this from CLUSTERA, it fails:
Could not find server 'CLUSTERA' in sys.servers.
And indeed, there is no CLUSTERA in sys.severs; the server_id=0 line of sys.servers is for SERVER1. But I don't want to use the actual SERVER1 name in my queries, of course -- that would miss the whole point of clustering!
So, how can I point to the cluster I'm in with a four-part name?