Like many startups and scaleups, Mentech didn’t have resources – budget or people – to waste; they needed to hit the ground running and deliver that efficient scaling quickly. But to support Mentech’s business over a longer term, Smits and his team wanted to build up their own cloud expertise. DoiT helped with both, quickly tackling the most pressing technology needs while providing tailored training sessions to build up Mentech team’s own knowledge base. DoiT’s close collaboration with AWS also offered Mentech a direct line to the teams building the products underlying Mentech’s infrastructure.
Migrating the data warehouse to Amazon S3 to increase reliability and reduce costs by 60%
There were a couple of known issues Smits and his team wanted to solve first. They were using an external SQL database, for example, which filled up too quickly, leading to increased costs.
“DoiT helped us to move from a monolithic and vertically scaling database that kept getting more expensive to a horizontally scaling solution that costs 60 % less,” says Sakis Kasampalis, Software Architect at Mentech. “We did it by migrating our large monolithic database to a combination of fine-grained AWS services: Amazon Timestream for the real-time data, Amazon S3 as the long-term storage, and Amazon Aurora for a short-lived subset of our data.”
Another issue was back pressure. Mentech’s system sometimes experienced trouble handling all incoming data from users’ wearables. With support from DoiT and AWS, the team identified the causes and took steps to improve the system, for example by storing metadata on Amazon RDS instead of Amazon S3 and by replacing HTTP 1.1 with HTTP/2. As a result, tickets raised and errors detected have dropped, showing a system reliability improved by at least 30%.
“We knew there was an issue with back pressure in our system, but we didn’t know where it came from or how to solve it,” says Smits. “It takes quite a lot of in-depth knowledge with AWS to find a solution for these more difficult problems, and that’s exactly what DoiT delivered.”
Finding infrastructure weaknesses with a Well-Architected Framework Review
After solving the system’s most pressing cost and reliability issues, it was time to really optimize the infrastructure for maximum efficiency. DoiT led Mentech through a comprehensive AWS Well-Architected Framework Review, a systematic and proprietary approach to compare a company’s AWS workloads against AWS best practices. With support from the DoiT team, Mentech reviewed the results and implemented the recommended changes. Two such changes were to add a backup system for its databases to avoid data losses in case of a crash, and to implement Multi-Factor Authentication to increase security. Both of these changes served Mentech to increase their uptime and strengthen the security posture of their infrastructure.
“The Well-Architected Framework Review that DoiT led helped us to really pinpoint the weaknesses in our cloud infrastructure, and then to systematically improve the scalability and general reliability of our systems. With these recommendations and others, we’re now confident that our architecture is sound, especially after solving our data warehouse and NAT gateway issues,” says Kasampalis. “Working with DoiT was an excellent experience for us, as they leveraged AWS technologies to enhance our cloud environment, increase productivity and reduce cloud spending by 10%.”