This content has been marked as final. Show 25 replies
My situation is I have a running 18.104.22.168 RAC cluster using fixed IP addresses resolved through a HOSTS file. I am using ASM with OCFS2 for the vote and config files. In order to update you have to update ASM first. During the third step in the process (using the runInstaller) it requires you to enter a SCAN name. The only things I have been able to find about it are that it allows a single name to address the entire cluster at the server level (it is not the same as the cluster name). This is different from a tnsnames entry allowing all of the instances to be reached by a single name (aultdb for aultdb1 and aultdb2 for example.) It specifically states that the SCAN name must be registered with the DNS or the cluster names server.
This is the first I have heard of being able to address multiple servers in a cluster with a single name, seems like that could be a bit confusing. I am sure if I blew away my current 22.214.171.124 install and reinstalled it would probably work but I wanted to keep the database the same exact one in the same layout and setup. Has anyone else tried an update of an existing 126.96.36.199 RAC using ASM and OCFS2 for the vote and config disks?
Based on the information here:
See INS-40922 it would appear that the name must be resolvable to an IP Address. You are not going to be able to address multiple anythings with a single IP Address.
Please open an SR and report back to us the resolution.
I am sending a note to the docs team that the doc, perhaps, need to address this.
Actually I am getting:
INS-40719: IP address configured for Single Client Access Name (SCAN): string is already assigned to another system.
Cause: The installer could not find the IP addresses you selected as SCAN addresses. This may be because they are assigned to another system, or the IP addresses are not listed in the DNS or hosts files as assigned to this domain name.
I have, in the hosts file:
assigned the cluster name as an alias to the public IP in hosts, same error
assigned the cluster name to a new IP address not assigned to a eth card, same error
assigned the the cluster name to a new IP address assigned to the same card as the public IP, same error
I will retry number three with a reboot.
Null doesn't work.
From the docs:
188.8.131.52.1 IP Address Requirements for Manual Configuration
The public and virtual IP addresses must be static addresses, configured before installation, and the virtual IP addresses for each node must not currently be in use. Oracle Clusterware manages private IP addresses in the private subnet on interfaces you identify as private during the installation interview.
Configure the following addresses:
•A public IP address for each node
•A virtual IP address for each node
•A single client access name (SCAN) configured on the domain name server (DNS) for Round Robin resolution to three addresses (recommended) or at least one address.
The single client access name (SCAN) is a hostname used to provide service access for clients to the cluster. Because the SCAN is associated with the cluster as a whole, rather than to a particular node, the SCAN makes it possible to add or remove nodes from the cluster without needing to reconfigure clients. It also adds location independence for the databases, so that client configuration does not have to depend on which nodes are running a particular database. Clients can continue to access the cluster in the same way as with previous releases, but Oracle recommends that clients accessing the cluster use the SCAN.
If you manually configure addresses, then Oracle strongly recommends that you use DNS resolution for SCAN VIPs. If you use the hosts file to resolve SCANs, then you must provide a hosts file entry for each SCAN address.
So much for not using DNS! I tried a hosts file entry for each SCAN address, it didn't work.
So much for RTFM...maybe I shoud tell them were to put the FM...
Edited by: mikerault on Sep 10, 2009 2:28 PM