An expensive solution is no solution

Start getting used to saving big, even while exploiting the full power of the public cloud even in production. Public Cloud Support & Marketing, including technical documentation is often geared to use you as guinea pigs or selling you the most services - not providing the best solutions.

The Real Cost of the Cloud

Resource	Typical Cost	Cost w/ Reactor	Savings
Relational Database	$2,500 per month	$750 per month	70%	$50,000 vs. $15,000 for 20 instances
Virtual Machine	$450 per month	$150 per month	67%	$90,000 vs. $30,000 for 200 VMs
MongoDB	M40 (4 vCPUs, 16GB RAM) $759.20 per month	M20 (2 vCPUs, 4GB RAM) $146 per month	81%	$37,960 vs. $7,300 for 50 clusters
Total per Month			71%	$177,969 vs. $52,300
Total per Year				$2,135,628 vs. 627,600

This table indicates savings based on typical resource usage in Production.

• When your relational database is only used as a data store, it can be smaller than you ever thought possible.
  • While a typical production database can be thousands of dollars per month, with Reactor your database spend can be well under a thousand dollars per instance, even in production.


• Use the smallest cloud resources like VMs with two cores, 8GB RAM and high-latency networks.
  • This means significant dollar savings when you spend less than $200 per month, per VM as opposed to more than $400 per month, per VM with the recommended sizes from public cloud.
  • These savings quickly add up when you scale up from 10 to hundreds of VMs with growth.


• Our hybrid database architecture means that you can store billions of records on the cloud at a pittance of the cost of conventional tech.
  • This means you would be spending a few hundreds of dollars per database cluster, rather than thousands.
  • You can start smaller than you ever thought possible, and spend significantly lesser, while being able to get superior performance and scalability for the same data sizes.


• Get used to fine-grained control of cloud resources based on actual usage rather than arbitrary recommendations which fill the coffers of already fat public cloud companies.
  • Start from the smallest cloud resources to begin with. Maximize savings using elastic, horizontal scaling of cloud resources.
  • In some cases, you can stay at the same size and only increase disk size to get better Disk I/O if high Disk I/O is the reason for scalability issues. This can be significantly cheaper than going to a higher tier of a VM instance (as an example).
  • In some cases, you can increase just the RAM, and stay at the same number of cores, when low system memory is found to be the reason for scalability issues.
  • Only increase the core count when sustained high CPU usage is found to be the reason for scalability issues.