A Comparison of Performance and Data Latency in Operational Cloud Database Systems
An important class of database applications must satisfy multiple critical performance requirements concurrently, including:
- High Volumes of Transactional Processing and Data Ingestion;
- High Volumes of Queries; and,
- High Consistency, including retrieving records immediately after insertion with very low latency.
For example, there are databases used to monitor, and rapidly respond to price changes in publicly traded securities, where there are billions of trades per day. These companies want to monitor their portfolios and market data in order to compute their exposure to risk and decide what to buy or sell and how much. If they can make these decisions before other traders, they will have a powerful competitive advantage that will make a big impact on their business. The new transactions must often be visible in the database within milliseconds of their occurrence in the public markets. Many other database applications have similar, near real time requirements in industrial, commercial and engineering applications.
In these applications, transactions and record structures are relatively simple, but there is nothing simple about satisfying the demanding requirements involved, in which high throughput and low data latency must be delivered consistently throughout the day.
CONCLUSION: On the basis of its independent research, WinterCorp recommends that companies with needs for low latency, high performance transactional-analytic data management software take a close look at InterSystems IRIS, which compares favorably to all alternatives tested on AWS on a single node and in 1- to 4-node clusters. Compared to alternatives, InterSystems IRIS exhibits significant advantages in query throughput, insert throughput, data latency and query efficiency without special tuning or configuration efforts.