1 Reply Latest reply on Jul 9, 2014 7:55 PM by bperry

    AF Failover Processing

    markshannon

      We have PI AF set to function as a primary and a failover(secondary) component ie a collective. Our primary is located on a virtual server and our failover is located on a second and separate virtual server. Recently, we encountered an event which triggered our failover component to become the primary. However, we found that while our seconday SQL Server DB (AF) had been getting backed up, the AF Databases and supporting components have not been migrated from our primary to secondary version.. our secondary version showed as the most current data from 3 months ago.

       

      I know that PI AF has a component option to refresh under the PI Collective tab. Do any of you know of a way to automate or emulate in an automated fashion the refresh feature of PI AF Collective?

        • Re: AF Failover Processing
          bperry

          When an AF collective is functioning properly, the database is actively replicated from primary to secondaries. No further action is needed. But in your case, something broke....

           

          You can check collective status in PI System Explorer. If it's not happy, then that's a good reason to call OSI tech support. They should be able to assist: if the formerly-primary SQL database is alive and good - it should be possible to delete the AF collective and reinitialize it, preserving the correct DB as primary and getting it replicating again to a secondary.