Hi, I think you can use one lock then read it through the servers using ssh. If the file not exists on prefered server then start a backup else do nothing. Cheers, Mufalani Enviado via iPhone Em 13/06/2012, às 23:30, Adam Musch <ahmusch@xxxxxxxxx> escreveu: > Put the job in the cron on all three servers. > Comment it out on the servers where it's only going to run if the > preferred server is unavailable. > It's up to you to uncomment it out if you want it to run on a > non-preferred server, and I'd probably be focusing on getting my dead > RAC node back up and running. > > On Wed, Jun 13, 2012 at 1:41 PM, Don Morse <dmorse2@xxxxxxxxx> wrote: >> To list:: I'm configuring rman backups for our first RAC deployment, ... >> Now that we have a 3-node cluster, I'm trying to decide how to schedule >> backups; one of the nodes will be the preferred server, others will run the >> rman job only if the others are not available. I have seen reference to >> using dbms_scheduler, but this would not work if we wanted to run an offline >> backup. I would appreciate any ideas. >> Thanks! >> >> Don Morse >> -- >> //www.freelists.org/webpage/oracle-l > -- > Adam Musch > ahmusch@xxxxxxxxx > -- > //www.freelists.org/webpage/oracle-l > > -- //www.freelists.org/webpage/oracle-l