michelangelus - Fotolia

News Stay informed about the latest enterprise technology news and product updates.

Translytical data platforms provide options to unify databases

Forrester identifies top vendors in the translytical database category and common elements that help make the unified database model well adapted for multiple use cases.

There's more than one type of data platform that organizations can choose for different use cases, including transactional, operational and analytical.

Then there's the translytical data platform, a model that combines all three database types into a unified instance.

A key challenge that translytical data platforms aim to solve is database sprawl. With a unified model, rather than having multiple databases to handle data and analytics, users can choose to have it all in one.

Using an operational or transactional database and then moving data to an analytics platform has long involved time and performance costs. With a translytical data platform, capabilities are built in to handle data efficiently for multiple applications.

The market for translytical data platforms is expanding, with multiple vendors all vying for a share. Oracle, SAP, Microsoft and IBM are all top ranked in the category in Forrester's fourth-quarter 2019 report on the translytical data platform market, released Oct. 23.

"Translytical databases vastly simplify business processes as one needs to maintain only one database, without the need for complex data movement, federation and reconciliation across multiple databases," said Tirthankar Lahiri, senior vice president, data and in-Memory technologies at Oracle. "As businesses now transform into real-time and digital enterprises, the need for translytical capabilities becomes even more acute."

For instance, Lahiri noted that financial transactions such as credit card charges or debit card withdrawals must be evaluated for fraud before they are allowed to be completed. This requires complex fraud scoring analysis to run within the boundaries of each financial transaction.

Also, telecommunications vendors must constantly optimize their networks based on real-time network congestion metrics -- metrics that are being continuously gathered and analyzed while an extremely high volume of cellular traffic is also being processed. This requires translytical or converged processing of network analytics, session establishment transactions, authorization and billing transactions, all within the same database.

Translytical data platform capabilities

Several characteristics and features help to define a leading translytical data platform. Among the key features is an in-memory capability for data processing.

In-memory is an important component in a translytical platform that ensures low-latency access to critical data used for various workloads, noted Forrester analyst Noel Yuhanna.

Translytical is a unified database that supports transactions, analytics, operational insights and other workloads in real-time without sacrificing transactional integrity, performance or scale.
Noel YuhannaAnalyst, Forrester

"Translytical is a unified database that supports transactions, analytics, operational insights and other workloads in real-time without sacrificing transactional integrity, performance or scale," Yuhanna said. "Historically, we have separated these workloads with an independent database because of performance reasons."

With memory prices continuing to fall and new Intel Optane persistent memory starting to attract customers, Yuhanna noted that users can now have hundreds of terabytes of in-memory for databases that changes the way users store, process and access data for applications and insights.

"Translytical is the next-generation database platform that leverages in-memory including DRAM, persistent memory and SSD, to support multiple workloads," he said.

SQL and ETL

Underpinning all data platforms are two core concepts, the ability to load data and the ability to query data. With traditional data platforms, SQL has long been the standard for queries. This standard holds for translytical data platforms as well.

"You can use SQL to access translytical, which is sufficient to support various workloads including operational, transactional and analytical," Yuhanna said. "But you can also use APIs to enable developers to build even more sophisticated applications using Java, Python, Perl, or other popular languages."

While SQL still has a place in the translytical data platform landscape, the same is not true for ETL (extract, transform and load). Yuhanna said no ETL is required with translytical, since it eliminates data movement. Traditionally, with a transactional database, an ETL job would move data to an operational database and from there to a data warehouse or analytical platform.

"Since all of these workloads are in a single database, ETL is not needed anymore," Yuhanna said. "This elimination of ETL is why translytical is a different type of database, and can deliver real-time analytics and operational insights, since you can run the queries as soon as the transactions take place."

The translytical data platform space will continue to evolve with vendors adding new capabilities, according to Forrester.

"We are likely to see more automation in the space, with automatic query optimization, further integration with Intel Optane persistent memory, and automated tiering of data across DRAM, flash and SSD," Yuhanna said. "We also expect translytical to become further cloud-enabled, with translytical-as-a-service offerings in the coming years."

Dig Deeper on Database management systems (DBMS)

Join the conversation

1 comment

Send me notifications when other members comment.

Please create a username to comment.

How does your organization use a translytical data platform?
Cancel

-ADS BY GOOGLE

SearchBusinessAnalytics

SearchAWS

SearchContentManagement

SearchOracle

SearchSAP

SearchSQLServer

Close