Re: detection of primary server failure

I should have been clearer. What I meant was that when the primary server goes down, is there some sort of alarm generated on the secondary server (with appropriate alarm code). If yes, then I can modify the alarm shell to convert the secondary server to a normal server. I remember Art mentioning it to me about some alarm which the secondary server gets when the primary goes down, but I lost all my emails in disk crash. Ravi "Esteban Casuscelli" <estebanc00@yahoo.com.ar> wrote in message news:8a8f14b9.0411290522.410b57ad@posting.google.com... > If you are looking for "the faster way" you should check if both > servers are up and running or online or whatever you want to do to > check if the engine is up. > > Then if the primary goes down you should check first if there are > shared memory segments or oninit processes still running there, if not > try first bringing the primary back up. If it fails then change you > secondary as primary and point your clients to the "new" primary. > > I think it is faster if you try first bringing you primary back up > rather than changing you secondary as primary and pointing your > clients to the new primary. And it s worth it to check what happened > on the primary before poiting your clients to the secondary server. > > hope this helps, > esteban.- > "rkusenet" <rkusenet@sympatico.ca> wrote in message news:<30tsubF3400s3U1@uni-berlin.de>... > > HDR > > > > IDS 921UC4 > > > > In this version DRAUTO has been disabled. Now it is only manual. > > How can secondary server detect the failure of primary server in the > > shortest possible time without involving a script to poll. I want the > > detection > > as fast as possible for minimum downtime so that the secondary switches > > itself > > to primary.