Judging Analytical Database Performance in the Cloud

Data-driven organizations rely on analytic databases—which are often in the cloud—to load, store, and analyze volumes of data at high speed to derive timely, actionable insights. However, data volumes within modern organizations' information ecosystems are rapidly expanding—placing significant performance demands on legacy architectures.

To gain the full benefit of their data for competitive advantage, businesses need modern scalable architectures and high levels of performance and reliability.

Recently, during a live DBTA webcast, William McKnight, president of McKnight Consulting Group, and Emma McGrattan, SVP of engineering at Actian, discussed the issues related to selection of an analytical database and how to evaluate database performance.

Performance is a very important aspect of a database selection for analytics. Cloud deployments are at an all-time high and poised to expand dramatically. Workloads are increasingly being moved to the cloud and new workloads are being considered for the cloud first and foremost, McKnight observed. Even if an organization’s users do not require high speed analytics yet, that should be the goal, he emphasized, because part of a data professional’s journey is to increase demand for data and its effective usage.

McKnight Consulting Group has also conducted a benchmark study on the performance of cloud-enabled, enterprise-ready, relationally-based, analytical-workload solutions. The benchmark results were insightful in revealing the query execution performance and some of the differentiators. McKnight shared his company’s evaluation of relational database platforms deployed using Amazon Web Services EC2. The evaluated databases were Actian Vector, Amazon Redshift, Microsoft SQL Server, Snowflake, and Cloudera Impala.

McGrattan provided a look under the hood of Vector’s X100 engine, explaining the key capabilties that help drive the database’s speed, including vectorized processing, taking advantage of in-chip cache, that it is a second-generation columnar database, uses smart compression, is able handle queries efficiently by eliminating unnecessary blocks of data from processing, and its multi-core parallelism.

To access the full 1-hour webcast with presentations by McKnight and McGrattan, audience questions, and the results of the McKnight Consulting Group  platform comparisons, go to