onbar on an inactive node?

Can I run onbar on an inactive node? We have a cluster env(ACTIVE/INACTIVE conf) and wants to run the onbar backup process on the inactive node? Thanks, Yogesh

What are you asking? IDS doesn't do traditional clustering (until IDS v11.10's SDS servers are available next month anyway). Are you talking about a 'cold' backup server on a different node of a cluster ready to start up if the primary server goes down and recover from the shared disks? Do you mean an HDR secondary? Do you mean an ER replicant that happens to be offline? What? Art S. Kagel ----- Original Message ----- From: Yogesh Nath <ids@iiug.org> At: 6/29 13:16:10 Can I run onbar on an inactive node? We have a cluster env(ACTIVE/INACTIVE conf) and wants to run the onbar backup process on the inactive node? Thanks, Yogesh ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

So as you said, it is a cold server on a different node of a cluster ready to start up if the primary server goes down. --Yogesh

Then no, you cannot take an archive from there. Two reasons: 1- IDS participates in archives, it does all of the reading of the data on disk and even flushed preimages of pages modified during the archive to the archiver (ontape or onbar) after each dbspace is completed. So without a running instance on the machine you can't archive it, and you cannot have another instance looking at the same disk as the primary (at least until you upgrade to IDS11.10 when it comes out in few weeks and can start an SDS secondary server on the cluster node. THEN you will be able to use the secondary for archiving, yes.) 2- If you COULD archive without a running instance on the machine, you still couldn't becuase the disk contents are inconsistent while the primary is online! IDS does ALL of it's operations in its buffer cache in memory writing log records to disk for recovery purposes. IDS only spools modified data pages to disk periodically, so the on-disk contents are always inconsistent as long as the engine's running. So, either take your archives on the primary server, or wait for IDS 11.10 and set up an SDS secondary on one or more cluster members and archive from there. Art ----- Original Message ----- From: Yogesh Nath <ids@iiug.org> At: 6/29 14:20:27 So as you said, it is a cold server on a different node of a cluster ready to start up if the primary server goes down. --Yogesh ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.