Cloud computing is opening doors for traditional software companies in ways some people might not expect. Sometimes it’s not about how to transition on-premise software to software as a service (SaaS), which has its own inherent issues for those contemplating the move. Rather, it’s about thinking how a given piece of software might work in a cloud computing environment. In fact, there is some software that has been designed to solve a particular set of needs in a traditional data center environment that can cross-over and work well in a cloud computing environment.
InMage, based in Santa Clara, California, is one of those software companies. Interestingly, network storage veteran Kumar Malavalli, who was the CTO at Brocade, founded the company back in 2002. Back then, his premise was to come up with a recovery solution that could protect application environments regardless of the underlying storage. Today, the company has over 500 customers and the product is sold entirely via a channel model.
Core Business: Application and Data Recovery
InMage provides a software-based application and data recovery platform that supports remote disaster recovery, continuous local backup, and application high availability at both remote and local sites. Because it uses disk-based data protection, reliability and recovery speed are significantly improved relative to legacy tape-based infrastructures.
Digging under the covers, some of the things that make the company unique in the disaster recovery (DR) space also make it a good fit for cloud computing environments, whether leveraging a public cloud, like Amazon or Terremark, or when used as part of a private cloud within the enterprise.
According to Eric Burgener, senior vice president of marketing, “InMage captures changes to data in real time as they occur and offers flexible recovery to any previous point in time, a feature that ensures fast, reliable recovery even when data is corrupted.” InMage uses continuous data protection (CDP) and asynchronous replication technologies to collect changes from one or more servers and sends them through an appliance, which can be local or remote, called the InMage CX. The CX can simultaneously provide the recovery foundation for multiple applications.
InMage uses a variety of storage capacity optimization technologies to protect very large data sets while using minimal bandwidth. As Burgener explains, “once an initial baseline is set it collects only data changes. As it moves data between source and target servers, it employs WAN acceleration and bandwidth shaping algorithms to minimize the amount of data that has to be sent to a given target to support recovery operations.” The result is a very low overhead solution that allows data to be recovered as soon as it is created, allowing customers to hit very stringent recovery point objective (RPO) and recovery time objectives (RTO). Each customer, depending on their needs or requirements, sets the targets for RPO and RTO in their environments.
Moreover, InMage’s architectural design flexibly accommodates Windows, Linux and Unix servers, physical and/or virtual servers, and heterogeneous storage and storage architectures (e.g. DAS, SAN, NAS, iSCSI and FC), while providing application-aware recovery solutions for the industry’s most popular enterprise applications, including Microsoft Exchange, SQL, SharePoint, Oracle, SAP, Blackberry Enterprise Server, and MySQL, among others.
Delivering Cloud-Based Infrastructure
Taken all together and thinking about how an enterprise or mid-market customer, or infrastructure as a service (IaaS) provider, might deploy InMage in conjunction with, or as part of, a cloud-based environment, is where things start to get interesting. The possibilities quickly open up.
For example, think about virtual server environments, where virtual machines (VMs) typically are run at 85-90% utilization rates. Conventional backup approaches that use host-based agents can easily take up 10 – 20% or more of a server during data protection operations, with the result that VM-based applications may take a huge hit. To Burgener, CDP is the perfect answer because of its low overhead. “CDP addresses a lot of data protection issues, but it is particularly appropriate for VM environments,” he says.
InMage’s combination of technologies provides simple, comprehensive recovery when it’s used as part of an in-house infrastructure. Interestingly, though, a lot of the benefits that make it a nice play there – such as keeping source and target data sets in sync over long distances without impacting production applications, the ability to recover both data and applications, very low overhead, and broad heterogeneous support – also make it a good play in cloud computing.
That’s good stuff, but what gets the team at InMage really excited about cloud computing are the providers that offer not just on-demand storage but on-demand servers as well. “Our integrated application failover and failback capabilities make it easy for customers to fail applications over onto VMs within the cloud infrastructure,” says Burgener. Cloud providers that can provision VMs in addition to storage capacity can offer a more complete set of services built around customer problem areas like disaster recovery, administration, and production. That means cloud providers could potentially offer everything from simple offsite backup to full-blown DR, adjunct test and development systems that do not hit capital budgets, and additional production resources accessible on demand to meet seasonal requirements.
To Burgener it’s this ability to do cloudbursting—the ability to “burst” into a cloud to access compute, network, and storage resources on demand —that is really going to make the cloud interesting to end-user customers and the providers that can leverage and deliver it. That’s why InMage is targeting cloud providers that can provision both servers and storage. “Provisioning storage in the cloud has interesting uses, but it provides a less comprehensive value proposition than something like cloudbursting which can bring application services into the mix,” said Burgener.
Cloudbursting provides a way for IaaS providers to differentiate themselves from the “storage services” crowd. The vertical integration enabled by cloudbursting may help cloud providers to attract more “storage services” business as well, since there is value to end users just knowing that their provider can offer cloudbursting if and when they might need it. “If their data is already available in the cloud, it makes it that much easier to start to think about how to leverage cloudbursting to meet additional recovery, administrative, and production requirements,” says Burgener.
Foundation for Public, Private and Hybrid Clouds
InMage’s technology is a good fit for public, private, and hybrid cloud environments. In fact, it’s in the area of private clouds and hybrid clouds that Burgener thinks cloud computing will find its biggest adoption over the next few years. Customers with InMage as part of their in-house infrastructure already have a leg up if they want to experiment with the cloud, since remote targets can be added (for data distribution and/or failover purposes) without having to touch their existing production servers.
Today, InMage is ramping up its cloud computing push with the announcement of its “cloud-optimized infrastructure”, which both describes its new offering and acts as a checklist for customers—end users and cloud providers—to gauge how well their data center infrastructure is optimized for cloud computing. The new offering includes flexible licensing, low overhead, non-disruptive scalability, enterprise multi-tenancy and support for heterogeneous environments. According to Burgener, InMage’s cloud-optimized infrastructure addresses three primary areas: disaster recovery and business continuity; test and development; and supporting spikes in demand. In turn, they are helping to push cloudbursting from concept to reality.
Of course, it shouldn’t come as too much of a stretch to find InMage pursuing the cloud computing market. What is surprising is that they didn’t build their solution with cloud computing in mind. Way back in 2002, like most people, Kumar Malavalli wasn’t thinking about the cloud. He was trying to find a way to provide a single solution that could protect application environments regardless of the type of storage they were built around. The fact that he did so and that it fits well within the cloud computing paradigm speaks to the overall vision of the man and the company he helped found.








