MongoDB Atlas provides Vainu Scalability, Security, and Reliability Using AWS

Vainu is an intelligent business directory service that collects B2B data about companies from a variety of public sources. It then sorts the data and makes it easily searchable. This provides its customers with a valuable information feed that can be integrated onto their CRM platforms or sent to other business applications using an API.
Founded in 2013 by three friends, Vainu was built on a tiny budget using an SQL database and a single on-premises server. Within 3 years the company was already generating more than $2 million in revenue, and had migrated its database to MongoDB. “Using a SQL database became a nightmare because of how often the schemas would keep changing,” says Kimmo Parviainen-Jalanko, head of engineering ops at Vainu. “We set up a MongoDB cluster using virtual machines and managed it ourselves.”

The company continued to grow quickly and managing its servers was becoming a burden for its employees. Vainu didn’t have a dedicated operations or infrastructure team so its employees would have to take on additional infrastructure maintenance tasks. “We were a startup with people doing system admin jobs in addition to their main duties,” says Parviainen-Jalanko. “The self-managed cluster couldn’t scale easily and we were always firefighting, everything was reactive."

About a year after making its first move to MongoDB, MongoDB Atlas was launched on AWS, and Vainu recognized that the fully managed cloud database could be used to transform its business.

It took about 7 years for Vainu to complete its migration to the cloud using AWS and MongoDB Atlas. Once its self-managed cluster was finally decommissioned in October 2023, the company made an immediate saving of €14,000 per month. “Our infrastructure costs for that cluster fell from €17,000 per month to €3,000 per month because we moved from an underutilized, outdated, unmanageable, and unscalable cluster to MongoDB Atlas on AWS,” says Parviainen-Jalanko.

Since Vainu first started using MongoDB Atlas, it had been migrating elements of its production environment from self-managed to fully-managed. “Over the years we had been moving stuff to Atlas but we couldn’t scale down the original cluster even though its usage was decreasing,” says Parviainen-Jalanko. “It was a big victory for us.”

Vainu containerized its applications using Amazon Elastic Kubernetes Service (Amazon EKS), a managed service to run Kubernetes in the AWS cloud and on-premises data centers. For general storage requirements it uses Amazon Simple Storage Service (Amazon S3). Vainu stores certain data for its larger customers in Amazon S3 to make it quick and easy to access. “Lots of our larger customers use the API in an asynchronous way,” says Parviainen-Jalanko. “So to reduce search times we provide them with a pre-signed URL to the location their data is stored. We have 90 million domains in our global database so without that it might take too long.”

Another important issue to be resolved was downtime caused by the manual management of domains. To overcome this Vainu used Amazon Route 53, a highly available and scalable DNS web service. “One source of downtime for us was when domains expired, which was really annoying,” says Parviainen-Jalanko. “We no longer use standalone registrars as it makes more sense to have all the domains auto-renew using Route 53.”

Migrating to MongoDB Atlas using AWS has significantly improved Vainu’s reliability, uptime, and security. Downtime was a serious issue with its previous infrastructure. “One of the biggest benefits of using MongoDB Atlas on AWS is the reliability and the cost visibility,” says Parviainen-Jalanko. “With Atlas, we can scale the clusters or we can easily spin up new clusters.” Through the AWS ISV Workload Migration Program, MongoDB and AWS simplify the migration process and reduce costs for customers migrating on-premises workloads to the cloud.

Vainu’s developers no longer spend time managing the infrastructure, which means they have more time to spend on improving the main applications and innovating. “This not only helps with scalability, it also helps with experimentation and innovation—our developers can easily spin up a test environment and play with something new.”

The costs associated with spinning up a test environment now are also completely transparent because unlike using a self-managed cluster, Vainu is only billed for the resources it uses. “Now, when a developer creates a new environment for a test project, we know exactly how much that test costs,” says Parviainen-Jalanko. “Previously it would either fall under the cost of the cluster, or it would require additional investment.”