Data Center Infrastructure and User Experience

Although SaaS providers are built for scale, cloud delivery involves a complex chain that starts in the data center.

06:57

Jan 03, 2017
1,353 Views

Although SaaS providers are built for scale, cloud delivery involves a complex chain that starts in the data center. Therefore, the economics of performance and user experience demand an agile infrastructure strategy sufficiently robust to scale as the user base grows.

In this video, Kaminario’s Vice President of Marketing, Josh Epstein, explains why data centers are the first step in ensuring great customer experience.

Transcript

Michael Krigsman: I’m Michael Krigsman, industry analyst and host of CXOTalk; and I’m talking with Josh Epstein, who is the Vice President of Marketing at Kaminario. Hey Josh, how are you?

Josh Epstein: Hi Michael, I’m great! How are you?

Michael Krigsman: I’m good! So Josh, many of Kaminario’s customers are SaaS providers. What are the data center challenges that these companies face?

Josh Epstein: You’re right that a majority of Kaminario’s customers are SaaS providers or similar on-demand application service providers. These companies are unique in that they’re built to scale, right? Their entire company, their entire business model is based on scale. Therefore their infrastructure has to scale. When they make decisions about storage, about product infrastructure, it’s always about scalability in mind. Second, it’s truly about they compete on user experience. The switching costs of a large enterprise actually switching from one SaaS provider to the next is really low. So they really have to consistently have user experiences: That means always introducing new functionalities. That means always having extremely high performance; so the infrastructure strategy has to match the performance and this agility of functionality. And finally, it’s all about cost. This is storage. Really if you think about the infrastructure of a SaaS company, it’s really the cost of goods sold in these types of companies. So, meaning a very cost-effective strategy is important, and really this predictability of cost efficiency throughout the life of the business as they grow, as they scale, as they evolve their platform.

Michael Krigsman: So given this key focus on cost, how do you recommend that SaaS providers manage this tension, and this challenge to deliver great user experience and at the same time, keep the cost extremely low?

Josh Epstein: Yeah, it’s really when you think of it as some might call the “economics of performance,” right? And it’s really easy, I guess, for a one-op product to go out, and do a little bit of analysis, and pick the best solution for that point in time - that most cost-efficient that delivers the kind of performance you need. But you really, if you try to take the long view, and try to understand how are your applications going to change as your user base is going to change, as more and more people are using more and more devices to get in; and then the application delivery itself becomes more and more complicated. Just thinking through how those things can change, and then thinking how the economics at scale work can really be economics continuing to deliver high performance as all of these new events come up, as these new scenarios, these new growth models start to play into that provider’s requirements.

Michael Krigsman: And presumably, the provider has to also be thinking about new devices, new sources of bandwidth consumption, and at the same time, as the load increases, maintaining that very quickened, responsive user experience.

Josh Epstein: Sure, and I would say that even more than just the network and bandwidth change, it’s really about functionality. I mean, these SaaS providers are really about introducing new functionality, they’re all very agile in the way that they respond to their own customer demands by introducing new functionalities. Now a lot of these new functionalities are based around real-time analytics or things like that. These actually put really complicated workflows on the infrastructure itself. So again, it’s about choosing infrastructure strategies that can be very agile; that can respond to these changes; that it can deliver consistently high performance [...] as the actual workflow has changed. And this is difficult for tradition infrastructure to do, and so that’s why this new wave of infrastructure providers, like Kaminario, are really focused on this element of technology.

Michael Krigsman: This agility that you just mentioned: What is the difference between the traditional suppliers of infrastructure, and modern infrastructure providers, and how do modern providers enable the SaaS vendors to provide this set of services?

Josh Epstein: When when we think about that in terms of agility, and really, you know, from a developer standpoint, it’s all about agile development; it’s all about being extremely responsive to end-user needs. We think about agility in a very similar way. It’s really the infrastructure has to be agile in a few different ways. Number one, it has to be able to agilely support growth. So again, you add new users, new devices, new requirements you need to respond. As said, you also need to be agile in the way of accommodating new functionality. [You could] layer on real-time analytics that always sort of becomes a very different workload on the traditional database application; very different from kind of classic OLTP database-type exercises when you are doing now real-time analytics. The ability to respond to that is important.

And then also, there’s a level of agility of being agile to adopt new technology  ─  new hardware technologies [and] new software technologies  ─  as they come down. So, the rate of evolution of advancement in the underlying technologies that make up storage, whether it’s processing tech, or network tech; or actual drive tech, or solid-state disk tech, is incredible. So it’s really important that the platforms that the SaaS vendors invest in can really be their platform over time, so as this new, safe flash drive becomes available, they will take advantage of the most cost-efficient, most high performance, most functional underlying technology that they can, and not walk into just a … They’re not locked into a scenario, they have to throw everything out and bring a whole new platform in, and really evolve perhaps.

Michael Krigsman: So there’s a high degree of flexibility. The last thing I want to ask you is: So you talk to many customers, and your most successful customers; how are they managing these challenges?

Josh Epstein: Now, that’s a great question. The most successful customers that we work with really look at storage, and really the entire datacenter as an application of their entire delivery strategy. If you think about it, a SaaS company is sort of different from a traditional software company. Their product is the functionality of the software that they deliver. They’re also the entire application stack behind that, right? They make some very strategic decisions. Now the smallest SaaS vendors might build out on say Amazon, or some sort of public cloud infrastructure. But as they reach scale, they start to realize that they can really have the strategic enablers and some real differentiation by optimizing their application delivery infrastructure, including storage. So they’re optimizing the performance for low cost for all the things that we talked about. So really, the most successful SaaS vendors are thinking about storage as an enabler, and therefore, they’re investing the time to effectively design in these infrastructure technologies into their overall application strategy.

Michael Krigsman: So storage becomes a core competency, in addition to application domain knowledge.

Josh Epstein: Yeah, it really is the entire application delivery stack that has to be a part of that overall technology strategy of the modern SaaS, the modern application journey of the company.

Michael Krigsman: Josh Epstein, VP of Marketing at Kaminario, thank you so much for taking the time.

Josh Epstein: Thanks, Michael!

Published Date: Jan 03, 2017

Author: Michael Krigsman

Episode ID: 407