The Race to Zero

How much should you pay to protect your mission-critical cloud workload against catastrophic failure? Given that the chance is almost zero that an AWS region fails, you should probably pay almost zero.

While there are many solutions providing disaster recovery automation for AWS, only one, our upcoming release of Thunder for EC2 Serverless, comes as close to zero as possible, while providing all of the features and extensibility you need to feel confident you have reliably protected your cloud workload against the unthinkable.

Disaster recovery protection is sometimes compared to insurance — you buy it hoping you don’t need it — so paying the lowest possible premium as a percentage of your monthly cloud spend should be everyone’s goal. This is particularly true for cost-sensitive small- and medium-sized business, or cloud consultants recommending solutions for their clients moving to the cloud.

Consider a small business hosting 10 modest-sized Linux-based applications in us-east-2; the AWS cost calculator might estimate a yearly cost to run these applications, as well as replicate 1 GB per day for each application across regions for DR protection, at roughly $2300 per year:

Disaster recovery protection involves provisioning offline duplicate instances in a remote region, driving the periodic snapshot replication, and testing the DR instances regularly. To achieve this you might consider purchasing an automation solution rather than spending your valuable time manually performing the operations yourself. But it only makes sense if the premium is reasonable: it might be difficult to justify, say, 10% of your yearly cloud spend on insurance if the risk is very low. Alternatively, you might consider completely going without protection and hoping nothing ever happens.

Our goal at Thunder Technologies is to hit the sweet spot of providing robust disaster recovery protection at the lowest possible price so that you no longer have to choose between breaking the bank buying insurance or going without. Thunder for EC2 Serverless transitioned our SaaS-based logic to AWS Lambda so that execution costs are the lowest possible. Given the sample 10-instance environment below, based on our tests using Lambda’s free tier and our 384 MB function, the yearly cost is calculated as follows:

That’s $27 per year to operate, just a 1% premium on even the most modest yearly spend.

This does not include the cost of the software itself, but because it is a function and not a SaaS instance, we can also be flexible on pricing. We can skip the monthly subscription fee and charge a one-time permanent ownership — at a price reasonable to end users but enough to sustain our business. And frankly, for disaster recovery where you might never need to use it, a monthly subscription in perpetuity might be hard to justify. Our goal in pricing is to keep the premium at or below 1% for small business — the only price we think is reasonable for cloud disaster recovery insurance.

Because we are a small company — writing software exclusively for the cloud requires little in the way of corporate overhead — merely capturing a small percentage of the gigantic AWS customer base is sufficient to be viable.

Because the cloud allows anyone to develop software for it with a very low overhead, the cost of enterprise solutions should continue to go down and down as vendors like us find more ways to wring out efficiency. In the race toward zero for disaster recovery, we are definitely in the lead and quite possibly near the finish line.

To make the cost even lower we are offering a discount on the purchase price to the first 5 beta customers. To sign up, or for more information, please contact us at info@thundertech.io