Redstor: Building a data management suite in the cloud

Q&A: Redstor has moved from backup into disaster recovery and archiving in the cloud.

Redstor is a data management software as a service business, offering backup, disaster recovery, and archiving in the cloud.

Its customers include the law firms Travers Smith, Hill Dickinson, and Nelson, the service management company Vivantion, as well as Capita, and Maxell. According to the company, to date, it has some 40,000 customers, ranging from sole traders to large international companies.

ZDNet talked to CEO Paul Evans and technical services director Thomas Campbell to find out more.

Redstor CEO Paul Evans.

Evans: Redstor was set up in 1998. In those days, we sold big solutions to enterprise customers, and our work went from putting in the first storage area networks (SANs) to developing high-availability systems, disk systems, backup and recovery systems, and so on.

Then we moved the business into managed services, which meant handling the backup and storage administration for our customers. But we had always had this vision that you would one day be able to manage your data like you manage your electricity, what we call 'data on tap'.

We wanted to build our own big storage platform and backup platform, and that was when we came across a company called Attix 5. It was 2004. They had this great piece of software which was designed for the cloud before people were calling it the cloud. They were backing-up to disk and it was multi-tenanted.

We then evolved the business into this big cloud backup platform and that grew really nicely. As the business evolved further, we were working with partners and, from there, we came across the education sector where we found a supplier looking to offer an online, backup service to schools.

That was version two of Redstor.

Now we had the intellectual property (IP) and we had realized pretty damn quickly that the IP was really good and, while the focus had been on backup, we now realized that there was a lot more to it than just backup.

Backup sits right in the middle of all the data flows. With backup, you can put an agent in every machine that you are looking to backup, and that makes it a single, central repository of data.

We had DR capabilities as well -- the ability to spin up a server in the event of a disaster. We had two-in-one protection.

Now, we want to talk about archiving. We are moving away from just back up to building a complete data management story.

We feel that a lot of enterprise customers are facing big challenges with their data. They've got lots of people working on it. They've got lots of tools. But we felt that there wasn't a complete data management tool which did it for everybody, was easy to use, affordable and easy to understand.

That's what we're building -- a complete data management suite for everyone. When we bought Attix 5, the first thing we did was completely rebuild the software and we called that Redstor Pro ESE.

The software's built in a modular way. Rather than having fixed releases, we went with releasing regular Agile releases. We made all the software modular with no technical debt whatsoever. It's all fresh code.

So where did the idea for the software come from?

Evans: I would like to tell you that it was all a cunning plan, but it actually came about by accident.

Someone came up and knocked on our door in Reading and said: "We're Attix 5, we've got this great software, it backs up to disk, it's multi-tenanted and it's priced in such a way that it's affordable."

Since we had the right price, we were able to create an online backup service that was priced in such a way that you could economically deliver it as a service. The other backup services that existed at that time were too expensive to provide an affordable service.

That's how we came across Attix 5 and we have always had this close relationship with them. Then, we wanted more control over the direction of travel the software was taking, so in 2015 we bought them.

One reason was that at that time our customers were telling us that they wanted faster releases so that they could get more functionality, more quickly, and so we moved onto a monthly release cycle.

Because we had our cloud platform, we got massive amounts of feedback on how to improve the product. Our customers really liked that, but they also liked that they were in control of the whole stack from software development through to the platform, through to offer really good customer support.

Because the customers are in the whole stack, their feedback gets through quickly to us and equally quickly back to them.

We have gone from being a little company based down in Reading to becoming an international software company focused on data management. We have customers in South Africa, North America, the Middle East, Australasia, and we're growing.

What's your unique point of difference?

This all-in-one approach is one of the key things that we have been doing. We are the backup and DR which was two-in-one, and now we are adding more functionality with archive, which is three-in-one.

We are doing things like offering a companion to archiving which can work out what data can be archived. That's the start of our data insight journey and that makes it four-in-one.

We are going to be bringing out a search and discovery app. Now, the great thing about backup is that you have always pulled your data back to a storage platform, a single repository. What better place to start your global search and discover that a single repository?

This is a really interesting area for us -- the ability of our customers to search all of their data in one place and/or to identify where that data came from. Because if you were to put the restore onto, theoretically, all of the data that was to be backed up then all of those individual data stores would be in that single repository.

What sort of interface are you going to have in this repository? Is it going to be highly technical or simple to use?

Campbell: We are trying to make it as easy to use as possible. Typically, it will take about 30 minutes to show someone how to do a restore. And then as far as all the machines are concerned and the connections between them, you can manage it all through a single interface.

It's a long way from the experience where you needed a two-week training course. Instead of leaving people to try to learn and manage all the intricacies of it, we have tried to make it really simple.

But are you going to leave the choices up to the customer because some will want everything in one place but some will want it diversified?

You're right, but for our standard offering, we will always keep two models of the data for our own. If they are a UK customer they can send the data to our datacentre in Slough and it gets replicated to our center in Reading.

We have lots and lots of options around keeping the data locally and they have the option to keep the data locally themselves and options around being able to do restore more quickly.

We've also done loads of work to make the experience of moving the data to the cloud platform as fast and efficient as possible -- one of the major things we've done is to use some functionality within the data.

What we do is make it appear that the data is recovered by using sparse files. We had a customer who had a RAID set failure and lost terabytes of data. Now what they can do is drop that entire dataset to disk. It takes a couple of minutes to create the sparse files on disk and immediately, from that point, you can start accessing those files on disk.

Instead of waiting, as you used to have to, for terabytes of data to come over from disk, you can now, within a couple of minutes, get things up and running again. And that's all happening over the internet.

SOURCE BY