Virtualization: A Promising and Justifiable Investment
Exclusive Q&A with Bala Murugan, Chief Architect, eG Innovations
By: Jeremy Geelan
Jan. 23, 2009 06:00 AM
"There is a shift in focus and it is from technologies that enable virtualization to technologies that manage virtualization," says Bala Murugan, Chief Architect at eG Innovations, in this Exclusive Q&A with SYS-CON's Virtualization Journal. Overall, Murugan maintains, virtualization is "a promising and justifiable investment, particularly in the current economic downturn."
Virtualization Journal: Do you agree with the view that Virtualization is one of the most promising technology investments in the current economic downturn?
Bala Murugan: Virtualization, when done right, has been proven to provide significant reductions in direct cost. It also helps your indirect cost by improving your IT’s performance, reliability and capacity management. So yes, I would say that it is a promising and justifiable investment, particularly in the current economic downturn.
Murugan: On the contrary, it is more in reference to the evolution of the Virtualization industry. Virtualization 1.0 was a revelation; it introduced virtualization to the world, proved its power and showed everyone how much they can benefit from it. Virtualization 2.0 – which is already here - is about accepting Virtualization as reality and moving on to how to do it right. How to get the most out of it. Essentially, there is a shift in focus and it is from technologies that enable virtualization to technologies that manage virtualization.
To be successful in Virtualization 2.0, organizations have to focus around technology that helps them manage their virtualization deployments better. Being a monitoring technology provider, we understand the complexities of monitoring in Virtualization 2.0 and are well positioned to help these companies realize the full potential of their virtualized infrastructures.
Murugan: We view Virtualization 2.0 as an evolution (next phase) – not as a radical revamp of current virtualization deployments. In Virtualization 2.0, the focus is on how to make virtualization deployments more cost-effective and how to gain maximum benefits. So this will actually make virtualization a mandatory technology for most organizations that are dealing with tight budgets in the economic slow-down.
Murugan: We live in an age of diverse infrastructures. Even before virtualization, the success of n-tier architectures and open systems made it impossible for one to have a homogenous environment. Data centers today comprise diverse technologies that have to co-exist and to in deliver IT services. Virtualization has taken this another step on the evolutionary road, now we are talking about adding a couple of more tiers to the n-tier apps by separating the hardware from the OS. At this juncture, we believe that interoperability is not a “nice to have”. It is a “must have.”
In terms of barriers, the ones that still exist are mostly technological, that people are working to overcome. In principle, I believe everyone agrees interoperability is a must have. Not only do they have to deal with a mix of virtual and non-virtual infrastructures, but also different types of virtualization from different vendors. They key we found is to be able to provide a unified consistent view across this diverse landscape, which makes management that much easier for the end-user.
Murugan: History has shown that Microsoft can be a significant threat in any endeavor it puts its mind to. They will have good technology and resort to their favorite ploy; their licensing model, and make Virtualization more of a commodity than it already is.
VMware itself has recognized that the hypervisor is no longer going to be the differentiator and that technologies that enable the effective use of virtualization (e.g., manageability), new application deployment models (like virtual desktops), etc. will be a key to retaining their leadership position.
Competition in this space can only be good – innovation will be faster and certainly there is room for multiple vendors in this fast growing market.
A: eG Innovations was founded by Srinivas Ramanathan, who also is our president and CEO. Prior to eG, he was a research scientist at HP and the chief architect of Firehunter, an ISP performance monitoring solution. His years at HP gave him a ringside seat to real pain points that customers have with monitoring their environments and monitoring tools themselves. In 2000, he left HP to build the proverbial “better mousetrap,” and assembled a strong team, including myself, to take this concept from the ground up. That was the genesis of eG Innovations.
Our focus was on monitoring n-tier architectures by looking at them as business services as opposed to a collection of servers, networks and applications. Our key benefit to the customer was our ability to proactively identify to the right problem, the true root cause, of poor performance in their IT infrastructures. As a result, customers spent less time firefighting and finger pointing, and more time improving their overall service levels. It took a couple of years to roll-out the finished product, and we got VC funding from Singapore. Then we opened up the US market in 2002 and found a receptive audience for the technology. We quickly became the premier Citrix monitoring solution, which had all the classic n-tier architecture issues. We won many awards and saw the company grow across the globe.
We saw the opportunity in the virtualization space quite early and started working with early virtualization adopters to better understand their needs and to strengthen our technology. Our mastery in thin-client computing and shared access technologies (Citrix, Microsoft Terminal Services, etc.) helped because a Virtualization ecosystem (one box – multiple OSs) is similar to a Citrix ecosystem (one OS – multiple users). More awards later, we are now recognized as one of the industry leaders in the Virtualization monitoring space, with support for different virtualization platforms including VMware, Citrix Xen, Solaris Containers/LDOMs and more.
Murugan: The biggest single pain point is probably problem isolation. When there is a problem in your n-tier IT infrastructure, it is usually pretty hard to distinguish between the true root cause and the effects. With systems being interdependent, a single problem generally causes a ripple effect that flows through the entire environment, leading you to chase effects as opposed to pinpointing the root cause. In simple terms, this means you are wasting valuable IT resources in fire-fighting mode fixing effects, which leads to finger pointing inside the organization. Meanwhile, your customers are still facing the problem. Virtualization only increases the complexity of your n-tier IT delivery, which makes problem isolation even more difficult.
Another key pain point that we see customers face is lack of visibility into their IT infrastructures. Even though it sounds simple enough, more often than not customers today don’t have total visibility into what is going on within their virtualized infrastructures. When you are managing a virtualized environment you definitely need answers to questions like; “How many guests are they running?” “How many guests are just consuming resources without being used?” “Where are the bottlenecks in the environment?” “Where do you stand on capacity?” “How do applications running inside VMs compare to ones running on physical servers?” “Is VMotion happening? If yes, why?” and so on. When it comes to virtual environments, what you don’t know can hurt you badly.
Another common problem is the classic disconnect between business services and the IT infrastructure. For example, business users say they can’t process orders or things are too slow. The IT side says servers are running fine on CPU. Both of them are right in their own perspective, but they are not on the same page, not even on the same book. This comes from the traditional IT view of looking at boxes and servers as opposed to the actual quality of services being delivered.
Murugan: There are many, but a classic one was when we got called in by a customer who was deploying a new project with Citrix technologies in a heterogeneous infrastructure with physical and virtual servers. Their new service was not taking off. Users were complaining about severe slowdowns and they had already spent weeks on this problem with no results. Before they came to us, they had changed the server hardware, the application software, the client terminals and software, all to no effect. Within a couple of days of getting involved, we were able to pin-point the source of the problem – network packet retransmissions between servers -- due to some issues with the way network teaming had been set up. We had been working with the application and server teams, and these teams had no visibility into the network. All they had to go by was what the network team was telling them. Hence, they assumed when a problem happened that it was a server or application issue, and spent weeks chasing this. Without any kind of instrumentation on the network, our eG Enterprise solution was able to determine that the root cause of the problem was in the network, not in the VMs, Citrix or other applications. This was a classic case of having to work with limited visibility into some domains, working with different silos of the infrastructure, and yet being able to effectively troubleshoot problems. In the end, it took us just a minutes to review the collected metrics to identify the root cause. Even after hundreds of customer installations, this remains a great example of a customer success.
Another very good example was a large financial institution where our technologies have delivered immense value. Before we got involved, they were very silo-based in their day to day firefighting and operations. We helped them streamline their operations, providing the helpdesk with end-to-end visibility into key business services. s a result, when a problem occurs, the helpdesk knows exactly which expert to call to resolve a problem. This produced significant improvement in service uptime, and more effective use of their operations staff.
Murugan: VDI and its various technology cousins are definitely here to say. The idea of a centralized desktop with the power of a localized desktop is extremely attractive. Some of the largest implementations have been VDI related. Currently we are seeing Fortune 100 companies leading the way on this and I believe it will be common place soon even in mid-size companies. As a technology, it has not yet fully matured, but once it does we see it as becoming a much bigger market than server-based virtualization initiatives. It may become the de-facto desktop platform in near future.
Murugan: I believe infrastructure has always been on the agenda of C-level execs, but with the success of virtualization there are definitely more conversations at the C-level about how to do this right.
Murugan: The platform independence provided by Java was one of the key drivers that enabled a slew of web-facing service-oriented applications in the last decade. Java and its sister technologies remain one of the backbone technologies of the web-based applications.
Reader Feedback: Page 1 of 1
Latest Cloud Developer Stories
Subscribe to the World's Most Powerful Newsletters
Subscribe to Our Rss Feeds & Get Your SYS-CON News Live!
SYS-CON Featured Whitepapers
Most Read This Week