Keeping a clean OS gold disk, standard web/app server, processing server, full text server. Load balanced on F5 with GSLB for web/app servers. Same for full text servers. Processing servers are identical except for the service definitions (timer, dip, scheduler, automation).
Has anyone successfully done multiple V2V with a master image, adjusting machine names, addresses, network connection points, security certs etc with a script and successfully adding capacity in a managed, cost effective way? We will have an upgrade every year plus three enhancements releases. This also almost ensures you won't see failed-over servers behave differently with respect to their outputs as mapped by GSLB.
Cloud services are certainly appealing, but we need to be super agile and customer engaged during our high growth phase.
Look forward to hearing some feedback and dialog.
