I think the issue with this is that you can't power on the original VM as it'll conflict with the failed over version in Azure. Double-take used to get around it by basically doing a system state restore over the top of the original server with a backup from the target (essentially swapping the identity of the VMs over) so then the original VM could be on the network receive replication updates. Maybe Microsoft could achieve this by giving us a Windows PE image we could boot off USB or something which would run as the target?
Currently working to understand the requirements for scenarios. Will update the thread soon
Other tools I've worked with have all had the option to select a network and IP address for production failovers and a separate network and IP address for test failovers. This is a significant ommission from ASR.
4 votesDave Lee shared this idea ·
Agree that analysis of physical machines would be really useful. I'm sure something could be done to bring perfmon data into the Azure Migrate project somehow. However, a nice (hopefully!) quick win would be the ability to manually add a machine into the Migration Project, push the agent out for in depth discovery and allow us to include it is the Service Maps. I'm more interested in that than I am in getting rightsized recommendations for target IaaS VMs based on perf data, that I can adjust myself.