Small, Smaller and Smallest Have Become the New SLAs for Recovery Windows; Interview with Dell’s Michael Grant, Part 2
Small, smaller and smallest. Those three words pretty well describe the application and file recovery windows that organizations of all sizes must meet with growing regularity. The challenge is finding tools and solutions that enable them to satisfy these ever-shrinking recovery windows. In this second part of my interview series with Michael Grant, director of data protection product marketing for Dell’s systems and information management group, he elaborates upon how the latest features available in Dell’s data protection line enable organizations to meet the shrinking SLAs associated with these new recovery objectives.
Jerome: When organizations go to restore data or an application, restores can actually take longer to complete because the data is stored in a duplicated state and they have to re-hydrate the data. How does Dell Data Protection | Rapid Recovery manage to achieve these 15 minute to two hour recovery windows?
Michael: We see prospective customers challenged with these lengthy times in recovery as well. If you are moving data a long distance, particularly if you have deduplicated it, you have now added re-hydration and latency to the equation.
At the same time, their onsite server recovery service level agreements (SLAs) have gotten small. We have already seen a lot of mid-market customers turning to Rapid Recovery to deal with this challenge. What they are doing is building something of a hybrid environment. Now, long-term, they tell us in no uncertain terms that when they find the ways and the means to get all of their data protection off site, they would like to do that. Will they really do that? I don’t know. But that’s long term. In the short-term, they are focused on building these hybrid environments.
When I say building a hybrid environment, typically that means they run a Rapid Recovery media server on site, and keep a full repository there. Then they replicate to public or private cloud. As part of what Rapid Recovery does, it spawns a hot standby virtual machine (VM), which is always running and available.
It updates as frequently as you take snapshots of your environment, and then replicates it automatically. For users, that means they can recover on site within literally minutes. They can recover offsite depending upon the latency. It is deduplicated throughout. But they can also access that media server directly.
In the event they have an outage where recovery time would be too onerous, they can access the media server directly running on the data that is minutes to no more than a half hour old, and work there while the IT team takes time to decide how they want to restore and where they want to restore. This is how we bridge the two, so that you get the data back, get the application back, and get the workforce up and running even though you may not have yet completed your entire restore.
In the DR series we see something similar. Re-hydration and line latency seem to take a bit of time. We are watching a lot of customers at this point put one or sometimes several DRs on site and then replicate between the two. In that scenario, you are not dealing as much with the re-hydration and latency between them.
Right now, given current capacities, customers easily keep 90 to 120 days of data next to the end users and systems that need it, so they may restore at line speeds within their environment, versus having to get it from off site.
But if you talk about the future of data protection, the next big challenge I think all of us face is, “How do we effectively put stuff at a distance?” So think away from the CPU, away from the storage, away from the end users, but still in a way that we can retrieve it almost instantly. That will be the new challenge that DR-as-a-service, backup-as-a-service, anything-as-a-service, really, has to tackle and solve. It has to perform these tasks to the satisfaction of the admin whose job is on the line if that application does not get back online and running.
Right now, I see a growing interest from customers and partners in getting as much data off site as they can. We continue to work with MSPs and with other partners to take a look at how can we do this most effectively while offering the same very low service level agreements (SLAs). We want to do so without unnecessarily being bounded by legacy architectures that many people have inherited, are stuck with, and have to figure out a way to augment and manage.
Something that also ties into this is that we have put out a free edition of our Endpoint Recovery offering. As part of the the digital transformation many of these businesses are going through, there’s a lot of IP that is getting created on the laptop while on the train ride to work or on the airplane to the next destination.
Our Endpoint Recovery solution is designed with the same principles in mind as Rapid Recovery. It’s a snapshot based technology that can frequently snap your entire image. In today’s version, it’s user driven, so an end user is responsible for both backing up and then restoring his or her data. You can restore granularly or up to a full image if you choose.
Later this year, we are going to put a management interface on that. For firms that are interested in managing everything holistically, we will provide that type of interface. The free edition is available for folks to experiment with and also to give us feedback through our data collection process so that we can understand what they are doing there.
If you think about end point all the way out to cloud, it is not hard to see how we connect the dots. We have got to protect the end point. We have got to help you get out to the cloud. The only way to do that, frankly, is to do it in cooperation with our customers. Having them tell us what they need, rather than us telling them what they need. That’s the impetus behind both this release and other features we plan to release later this year.
In Part 1 of this series, Michael Grant summarizes some of the latest features available in Dell’s data protection line and why organizations are laser-focused on recovery like never before.
In Part 3 of this series, Michael Grant shares about some of the new development going on in the NetVault and vRanger product lines.