Cloud Computing and Data Storage: Handling the Chaos – Datamation

In an era that requires massive data storage necessitated by data analytics and even routine operations cloud storage is now an unquestioned foundation for many businesses.

But while cloud storage offers a big helping hand, it's not without its own complexities. Among the several: the data is stored remotely, but we need the data right now. Also: do we store all our data with one cloud company? How do we manage a multicloud strategy?

To discuss these issues, I spoke with Ellen Rubin, CEO and Co-Founder of ClearSky Data, a cloud storage-as-a-service company. A veteran tech entrepreneur, Rubin also co-founded CloudSwitch, a cloud software firm that was later acquired by Verizon.

Among many topics, Rubin and I discussed:

See below: transcribed highlights of my discussion with Ellen Rubin.

So what we're seeing today is that a typical enterprise that is not born-in-the-cloud, never-had-a-datacenter, but instead has a mix of more traditional legacy infrastructure, as well as maybe multiple cloud providers. [They are] dealing with this very extended and complex environment where they have things like VMware and databases, Oracle, SQL, and all things sitting in more traditional data centers. Which of course they're desperately trying to consolidate and get out of.

But theyre still buying more gear, making copies of the data, backing it up, having a disaster recovery site, all the things they used to do, even if they don't have as many [elements] as they used to.

So what you have is almost like an a la carte menu for the people who are in the business side: where things could be run, and what needs to happen to support the business use cases and you've got IT in the middle trying to handle and manage it.

And what we're finding is that really, nobody has thought about the fact that what you want is not to just deal with accessing [the data] and pulling things back from the cloud, which is really expensive. What you actually want is: you just want the data to be accessible wherever the computer's running.

One of the hidden things that's true is how dramatically things are changing in the networking world because of this distributed environment. The need to be able to connect across physical distances but still deal with latency as a problem was something that nobody seemed to be dealing with.

So what you see customers doing is they are running lines and they're starting to take advantage of much more modern types of approaches, not just AT&T and Verizon, but some of the more modern on-demand, more API enabled types of services. But they're still kind of stuck with the fact that latency is a problem that is not going to be solved by them. And they also are dealing with the fact that they're not expert in that.

You can't solve this problem without addressing last mile issues and the fact that the data that's sitting physically in a cloud somewhere far away may not be that great, if your compute needs to be elsewhere or it needs to be distributed.

One of the pleasures of being an entrepreneur, especially 'cause I focus on IT-related infrastructure things, is getting to meet with CIOs all the time, and that across many companies.

And I have to say that this is as challenging and trying a time as I have seen for a lot of the heads of IT and CIOs. Because certainly, the CIOs have shifted into much more of a business type of a role where they're really meant to focus not just on how do I reduce all my costs down to as little as possible, but I'm supposed to help promote the digital transformation that the executives are trying to accomplish and the business units are trying to be agile. And be able to roll things out much more real-time in this user-based, self-service, everything on demand environment.

So that puts both incredible pressure [on them]. But it also, at the same time, requires cost production and those are very hard things to do.

But in a way, the assumption is that this is all supposed to be cheaper, and so it's just a tremendous amount of pressure in terms of the decision makers. And I think that their feeling is that some of the things, they're really, really great at. But other things would be best to turn to an outsource provider of someone who can act as a trusted advisor or provide services to them. So we're a small part of that ecosystem.

It's not like keep it all, analyze it all because rarely is that so beneficial. But also it's really hard to do, right? If you're moving petabytes of data across lines, that's gonna be pretty challenging.

So I think that's what's different. But there are also is this explosion of Edge Data Centers. So we spend a lot of time with our partner Equinix as a major data center provider. And they're in all the major metros but there are also players that are trying to be yet even further to the edge in cell towers and micro data centers and sites that are really almost just regular buildings, where a lot of the data and analytic catalogs could be captured and processed.

And so what it means is, even though people are consolidating traditional data centers, there's an explosion of more locations where some sort of infrastructure has to exist but it has to be small.

This whole pendulum swings between: it's distributed, no, it's decentralized, back and forth, so we're in a big decentralization moment, where more and more things can be processed locally, decision making can happen.

Certainly, IoT is a very extreme example of it and is very impactful to certain industries but your plain old regular traditional enterprise decision maker is a little less in the IoT world maybe. And more in the trying to have more and more access to real-time data that's happening at in people's phones, at points of interaction and local stores, that kind of thing.

And I think it's a very exciting time, but what it does is, yet again, puts a lot of pressure on how much of that are you going to own and build and manage yourself, versus how much can you take advantage of what's out there in the ecosystem from technology providers? The cloud is certainly a huge part of it, but it's not the only part.

We seem to be in a very long cycle right now. We were talking about it's 10 years of cloud. And now we've got all this edge stuff going on, and we're just in the emerging few years of that taking off. So I can't say that in the next five years it will have become simpler because that probably is a little optimistic.

But what I can say is, I am sure that we will see another swing of companies getting much more standardized on a few different key types of products and vendors, for the different types of things they're doing. There'll start to be more consensus.

And I see it by industry vertical, financial services versus healthcare versus law firms that we service, where they have certain things that they do, for different types of work that have to get done. And they all talk to each other, and what starts to emerge as: that's the right way to do it.

So I think we'll have more of that, even, I don't know what to call it, the playbooks, best practices, pick your favorite word. But what it will mean is that there'll be a lot less complexity in terms of the numbers of things that are all being tried at the same time. So at least, that hopefully, will be a little bit easier.

Go here to read the rest:

Cloud Computing and Data Storage: Handling the Chaos - Datamation