Vision Solutions Double-Take Software Steps Into Cloud Computing With Amazon EC2 – Promises ‘Recovery as a Service’

please note – Double-Take is now part of Vision Solutions

Double-Take Software Cloud Computing Highlights

• Has selected Amazon’s EC2 as the first deployment platform and partner
• Offers real time replication at around $250/month for the first and $150/month for subsequent servers
• Focuses on ensuring recovery, rather than merely backing up data
• As an asynchronous technique with a recovery point, will not be used for transaction processing
• Is only available for x86 servers running Windows or Linux
• Will move to ‘Recovery as a Service’ provider status
• Will consider partnerships and more integrated solutions in future
• Is focusing on Large as well as Medium and Small Enterprises

I had an interesting talk with Peter Laudenslager of Double-Take Software about his company’s first steps into Cloud Computing which were announced on February 22nd 2010. The company is a specialist in real time replication, with software products such as:

• Availability (which replicates changes in real time to one or more servers at any location around the world)
• Backup (which offers Continuous Data Protection – CDP)
• Move (which allows for server migration between physical and virtual servers – including a mix of VMWare and HyperV-based systems
• Flex (which allows for desktop and server provisioning and management).

The company is a relatively small supplier with annual revenues of $96 million in its last financial year, which ended in December 2009 and based in Southborough, Mass., USA. It also has offices in Indianapolis, New Jersey, the UK, mainland Europe and Asia. Its customer base is large organisations, typically in Legal, Health, Finance and Federal Government sectors – all types of users who place a high value on data and availability according to Peter. The company is focused on enabling the recovery of servers and systems and recovery time objectives.

Double-Take’s first move into Cloud Computer uses Amazon’s EC². Peter evaluated other potential partners, but chose Amazon because it has a valid approach, is already used by his customers and has significant regional resources. A customer will need a contract with Amazon as well as renting software from Double-Take for a flat fee of $150 per production server. To set up a replication with Amazon you will need:

• A small instance in an image repository, which is currently priced by at 13C/hour – adding to $93 per month
• Storage disk space, priced at 10C per GB per month
• Bandwidth, which Peter admits is difficult to predict – not least because Amazon has temporarily made in-bound bandwidth free of charge (having previously charged 8-19C per IOP); he estimates that the costs will be around 20C to 25C per calendar month net of storage IOPs

In total a customer will pay around $250 a month for the first server replicated and $150 for each there after. I was impressed with the fact that this is not per se a Small and Medium Business (SBM) solution, as Amazon has the capacity to support thousands of servers.

In use Double-Take claims significant advantages over traditional high availability solutions. He admits that, as his offering is asynchronous, it is unlikely to used in transaction processing, since it always works with a restore point – even if that is minutes, rather than hours or days. However for most customers, having an asynchronous automatic fail-over will be more than adequate and will save potentially millions. There are advantages in ‘time to deploy’ as well: a customer can read about the solution at breakfast and have the system up and running by the afternoon’, according to Peter.

He also claims advantages over Cloud Computing backup ‘Storage as a Service’ offerings (from companies such as Mozy, Carbonite and iDrive ), because the replicated system allows the customer to recover the system, rather than just backup data. He argues that there are many difficulties in recovering complete systems, not least because databases and interconnections are difficult (and expensive) to restore.

His services offers virtual redundant systems waiting to be called up in an outage, rather than the expense of buying, deploying and managing physical servers.

Peter believes the current trend for companies such as Symantec and EMC to offer deduplication relates to their batch processing past and the legacy of tape-based systems: he argues that it is unnecessary if you do real time replication.

In future he believes he will be able to launch the company’s ‘Recovery as a Service’ offerings on other Hosting companies’ platforms, considering Microsoft Res (once the platform is more mature), Rackspace (once it expands geographically beyond its US operations) and a number of Asian companies (as his customers there don’t necessarily want to be replicating data across vast distances).

Some Conclusions – The Arrival Of ‘Recovery as a Service’

I was impressed with what Double-Take is up to. As with many smaller suppliers it has thought creatively about how to simplify its offerings and make them available over the Web. The current need to have contracts with the company and Amazon doesn’t put this offering in the ‘… as a Service’ category for me, but that is clearly where the company is headed. I can think of a couple of downsides to the approach:

• The company currently only covers x86-based servers running Windows or Linux operating systems, so you won’t be able to use them if your running Sparc, PowerPC, Itanium or zSystem processors
• As he admits himself, this form of online replication is inappropriate for transaction processing, and/or for those customers who absolutely need to recover instantly and can afford fully redundant fail-over systems

Overall, however I believe this is an innovative approach and that the company will win many new customers of all sizes. In fact it will be very interesting to see where the new service takes off first, since it is one of the few I’ve seen that doesn’t start first with smaller companies and consumers. You should take a look at Double-Take’s offering if you’re interested in saving IT costs by concentrating on having ‘just in place’, rather than a ‘just in case’ recovery solution.

How do you ensure the future recovery of your servers? How do you get beyond data to system recovery? Have you used Double-Take software’s offerings? Let me know by commenting on this post.