We have a cluster with production server installed on one node and
analytical server installed on the second one.

We encountered problems during execution jobs on production server - jobs
retrieving data from linked server (analytical server installed on the
second cluster node). From time to time (about half cases) executed job
failes with message:

SQL Server does not exist or access denied.

What is a bit queer there are no communication problems from analytical
server to production server.

I think about problems with clustering. What's more during retrieving data
in query analyzer I also don't have problems - maybe there is something
wrong with SQL server agent.

The only hint I have: we migrated our domain to Active Directory and after
cluster restart we couldn't start cluster network name service for
production server (doubled registry in DNS which was deleted and service
started).

Regards.

Try forcing your connection to use Named Pipes instead of TCPIP to connect to
the server. This will probably clear up your problem. I had a very similar
issue just a few weeks ago.

"Curious" wrote: