December 01, 2011, 4:13 PM — VirtualSharp hopes to soon let its customers set up disaster recovery plans that fail over from their private clouds to public clouds, the company announced this week at the CloudBeat conference in Redwood City, California.
VirtualSharp's software lets customers do disaster recovery on their private clouds and also easily simulate failovers -- as often as daily or even every few hours -- in order to make sure the disaster recovery process will work properly in a real outage. Currently, customers use a different internal data center or additional virtual machines for the failover.
On Wednesday, VirtualSharp announced features in its ReliableDR software that will let public cloud providers provide the disaster recovery services. Companies will be able to sign up for the service online and set a so-called service level agreement for the disaster recovery process. For instance, the customer can say that it wants to be able to recover an application in 30 minutes and run a disaster recovery test every six hours. If those parameters aren't met, the cloud service provider can send an email to the customer indicating what is causing the holdup.
"This means that when you have a threat to recovery it will be found long before it impedes or becomes an obstacle to recovery" in a real outage, Carlos Escapa, CEO for VirtualSharp, said.
For now, the product is compatible with clouds -- public and private -- using VMware hypervisors.
VirtualSharp expects to be able to announce a public cloud partner in the coming weeks and anticipates service availability in the first quarter.
Pricing will be dependent on the public cloud provider. Escapa hopes to eventually achieve a cost of US$100 per virtual machine in the public cloud per month. Current customers license the software and pay a maintenance fee based on the number of virtual machines they use.
VirtualSharp represents a new kind of disaster recovery process, and not just because it automates the process, Escapa said. A company may be doing data replication for disaster recovery, but that's not enough anymore because companies are increasingly running complex applications that may span data centers.
"I feel strongly that the focus is not so much on data protection. There's a shift to service protection, which is data plus apps plus configuration plus storage plus the virtual machine," he said. "All of these things have to click together to make sure a service can be restarted if the cloud gets blown away."