Microsoft has just announced the preview of “Microsoft Azure Site Recovery” and we are glad to see that we are making progress in our hybrid cloud, letting customers extend their datacenter solutions to the cloud to ensure business continuity.
First, what exactly is Azure Site Recovery?
In order to give you a better understanding, we start with Hyper-V and Hyper-V Replica.
Hyper-V Replica was first introduced in Windows Server 2012, directly integrated into the hypervisor.
It was an affordable, flexible and easy to manage disaster recovery solution that let you enable replication at the VM level. You could replicate VMs to stand-alone hosts, Hyper-V clusters and vice versa. The scenarios it was designed for was enterprise 2 enterprise and enterprise 2 service providers. What I personally liked the most was that it was both application – and storage agnostic, leading to the most flexible DR solution on the market. We have been working with several customers since then where we have used their tier 1 site as the primary site, replicating to site 2 that is using JBOD, SMB and everything that is cost-effective with the innovations in Windows Server and System Center.
To summary, Hyper-V Replica is an asynchronous DR solution made available for everyone.
Windows Server 2012 R2 got several new enhancements and can be configured so that you can even specify the copy frequency on your replication at the VM level. The default in 2012 was 5 minutes, but this can now be set to 15 minutes or even down to 30 seconds (!). We have the option to create additional recovery points, both for the crash consistent snapshots and the app consistent snapshots. And last but not least – we can extend the replication as well, meaning we can replicate from our DR site to an additional DR site.
At the same time as 2012 R2 was announced, we saw that a new service in Microsoft Azure was born.
“Windows Azure Hyper-V Recovery Manager” was introduced at the same time and utilized the cloud abstraction in System Center 2012 R2 – Virtual Machine Manager, so that you could get a simple and automated DR solution where all of the processes were orchestrated from the service in Azure, through the provider on VMM.
When creating a cloud in VMM, you could choose to send configuration data (metadata) from the cloud to the Recovery Vault in Azure, where VMM was registered. We could enable protection on the VMs during deployment in VMM so that they started on their initial replication to our DR site once the job has completed. In short, we only had to enable protection for our cloud and everything was automated.
In addition to all of this, Hyper-V Replica has built-in workflows for planned and unplanned failover, and test failover. The replication is bidirectional and let you reverse the replication during failover.
Microsoft Azure Site Recovery is the new name, and finally removes any confusions related to the Cloud.
We are finally able to use Microsoft Azure as our disaster recovery site now (enterprise 2 Azure), so that you can have your entire virtual infrastructure up and running in Azure in case of a disaster.
This is a major step in the right direction to simplify organizations approach to a hybrid cloud, and removes any requirement for maintaining a disaster recovery site on their own.
Going back to the flexible deployment model, it is also important to note that Hyper-V Replica supports alternative IP configuration for the protected virtual machines as well, so VMs that you failover to your DR site can be live and kicking on the network from the start. This is all handled by Azure Site Recovery and let you map VM networks in VMM to your preferred virtual networks in Azure. During failover, we will see that the virtual machines will pop up in the virtual machines view in the service management portal in Azure, letting you manage and access them just as you would expect.
This also leads to a new scenario.
Consider an organization that wants to move to the cloud. They can now use Azure Site Recovery to facilitate that process with minimal downtime, and migrate their workload from their private cloud to Microsoft Azure with Site Recovery.
We are also able to failback (of course) from Azure to our Private Cloud, which truly demonstrate the vision of the Cloud OS with one consistent platform.
For a more technical description of the solution, feel free to visit my tech blog at http://kristiannese.blogspot.no/2014/06/microsoft-azure-site-recovery.html
We would like to invite you to our next webinar where I will be covering Azure Site Recovery (oh yes, there will be demos as well!) so that you can get a better understanding of the solution and continue the discussion with us at Lumagate afterwards.
Our webinar is scheduled for Wednesday, 2th of July. 10 AM to 11.30 AM.
Sign up and join the conversation by following this link: https://lumagate.wufoo.eu/forms/webinar-microsoft-azure-site-recovery/