With the rise of cloud computing and the growing need for scalable infrastructure, selecting an appropriate database management system (DBMS) is crucial. This study evaluates the performance of MongoDB, a document-oriented database, and Apache Cassandra, a column family-oriented database, within a container-based distributed system. The evaluation focuses on each database’s approach scalability and resource consumption under varying workloads. To simulate a production environment, a distributed setup using Docker containers was implemented, with database replicas ranging from one to three. The performance was measured using the Yahoo! Cloud Serving Benchmark (YCSB), testing workloads with two, four, six, eight, and ten thousand requests per second. The results demonstrate that MongoDB excels in read-heavy operations, offering superior scalability, while Apache Cassandra performs more efficiently in write-heavy workloads, though at the cost of increased resource consumption and reduced scalability. This methodology proves effective in assessing both non-relational DBMSs.