CBR talks to Greg Rusu, general manager at Zunicore, PEER 1 Hosting’s new public cloud division
Let’s start by looking at your background.
I joined PEER 1 in August 2011 to start Zunicore out of San Antonio in Texas. We started with a small team at first and went from start-up to launch in about 120 days.
Prior to this I spent four years with a systems management software company, and before that I was with Dell and AMD. I’ve always been focused on the launching of businesses and divisions and bringing them to market, which is what I was brought in to do here.
Why PEER 1 did feel the need to enter the public cloud space, and why did you go with a different brand – Zunicore – rather than calling it something like PEER 1 Cloud?
It was a natural step from Peer 1 Hosting and our customers were requesting it.
In terms of why we’ve branded it differently, there are a couple of reasons. First, we just wanted to try it and see how it works with different approaches to billing, licensing and customer service and so on. Second we also wanted to innovate without confusing customers about how it all fits together. We’re also going after a slightly different sector of customers.
However we do plan to align the two more closely going forward. When we first launched it turned out that many thought we were a partner of PEER 1 Hosting rather than a division of it. We want to integrate PEER 1 and Zunicore more in terms of capability as well as the brand. We should see much more of that over the next year, although some elements have already started.
So is Zunicore initially being pushed at PEER 1 customers?
We did start pushing it that way but we’ve signed up a lot of new customers, which was a bit of a surprise. There has been some crossover but not as much as we were expecting. We’ve so far signed up a number of SMBs, government agencies and web agencies.
Security worries are often accused of holding back cloud adoption. Where do you stand on that?
With this sort of tech you can spin up VMs for not very much money and there is tremendous power connected to the Internet. This makes it a honey pot for folks to come in with fake credit cards or to try and launch a DDoS. There is nothing new there for us, we have monitoring systems in place for this sort of thing.
Soon after our launch we noticed attacks were becoming more frequent and strong – in fact we saw it in our activity as soon as we launched. So we held back [from really publicising the platform] and updated our automated processes for dealing with this. After a month or so was saw it level off in terms of type and volume. We were able to ID the attack vectors and now that we have a good idea of the types of attack we can filter out specific instances.
The cloud hosting space is very crowded now. What do you do that your competitors do not?
What we offer is the concept of a "resource pool". This means the customer can choose exactly how much CPU, memory and storage they want and customise it to their needs. This is different from the small, medium or large VM platform that the likes of Amazon offer. That way means you often end up with lots of resource you don’t need. We offer a true pay for what you use model.
There are a few other features as well: We only charge for outbound traffic and not for traffic between VMs, as Amazon does; we don’t have a different price depending on your location; we offer 24/7 support at no extra cost. We also use persistent storage area network (SAN) storage instead of local hard drives. This makes it more efficient.
We pitch Zunicore as a professional cloud, we want as many enterprise class features as possible at no extra cost. We want our customers to just worry about their business. The competition is more geared towards the mass market, we’re slightly higher quality while remaining price competitive.
You mentioned the idea of only paying for what you use – what about the issue of scaling up or down as demand changes?
Zunicore offers auto scaling and load balancing. It is all monitored for load and if it increases we’ll see that. If it gets too high for 5, 10, 20 minutes or however long the customer wants we can automatically add memory or CPU. It can then be removed if and when the load drops. It can also be set up to automatically kick-in if the customer knows when traffic may peak, such as at Christmas time or during a sale if it’s an online retailer.