buchachon - Fotolia

MarkLogic Data Hub Service aims to ease cloud use of NoSQL DBMS

MarkLogic rolled out a cloud-service version of its NoSQL database management system, a move designed to make the technology more cost-effective for cloud users.

MarkLogic Corp. launched a hosted cloud service to make it easier and less costly for users to deploy its namesake NoSQL database in the cloud -- something that most of the company's existing customers have yet to do.

The MarkLogic Data Hub Service offers users a serverless environment, with MarkLogic handling system configuration, resource provisioning, database security and other management tasks. Pricing is usage-based under a capacity-unit model that lets users accumulate credits for processing capacity when resource needs are low and then automatically ramp up database throughput when demand spikes.

Initially available in the AWS cloud, the new service is built on top of MarkLogic's Data Hub Framework architecture, which makes the MarkLogic database a platform for integrating different types of operational data for transactional and analytics uses. The Data Hub technology is aimed primarily at large organizations that need to pull together diverse data sets from a variety of source systems.

MarkLogic users previously could set up and run their own cloud systems. But the MarkLogic Data Hub Service's elastic scalability and usage-based pricing provides "a modern-day cloud approach," said Adam Ronthal, a Gartner analyst. That could make MarkLogic a more appealing alternative to the Amazon Redshift cloud data warehouse and other database technologies on AWS, according to Ronthal.

MarkLogic is very good at what they do, but not everybody understands what they do.
Adam Ronthalanalyst at Gartner

But MarkLogic has some work to do to reach new users, he added.

"We joke about MarkLogic that it's the most interesting company no one's ever heard of," Ronthal said. "But like all jokes, that one has some basis in reality. MarkLogic is very good at what they do, but not everybody understands what they do."

MarkLogic's focus: Decreasing data friction

From Ronthal's standpoint, MarkLogic's biggest selling point is its ability to support what he described as "frictionless data integration" in complex IT environments. The database includes automated features that can be used to create curated data sets for downstream uses, a capability that positions the MarkLogic Data Hub Service to function as a data lake or logical data warehouse platform, Ronthal said.

For now, a large majority of MarkLogic users still run the database software in on-premises systems, said Ken Krupa, CTO at the vendor, based in San Carlos, Calif. Cloud users make up only "a few percentage points" of the MarkLogic customer base, Krupa said. He added, though, that the company hopes the MarkLogic Data Hub Service will give users more reasons to embrace the cloud.

MarkLogic is a multimodel database that supports XML and JSON documents at its core, but it also incorporates graph technology and handles relational data. In its Data Hub format, the NoSQL software is typically used to integrate data from systems in different business units to create consistent sets of enterprise data, which can then be analyzed or fed back to business applications, Krupa said.

MarkLogic Data Hub Service setup screen
The MarkLogic Data Hub Service uses a capacity-unit metric with baseline and burst processing levels based on an hourly cost.

Cloud service lets users burst capacity limits

The MarkLogic Data Hub Service's pricing model is based on a processing capacity metric that combines a set amount of CPU, memory and network resources. Users pay for a baseline level of the MarkLogic Capacity Units (MCUs) at an hourly rate that's charged whether the capacity is used or not. However, MCUs that aren't used each hour are credited to a user's account and can be tapped at a "burst capacity" level when more processing resources are needed, according to MarkLogic.

For example, the base price of $4 per hour includes a baseline level of 32 MCUs and lets users run up to 384 MCUs if they have that many available. The baseline level scales up to 512 MCUs for $64 per hour; users also pay a monthly fee of 10 cents for each gigabyte of data stored in the cloud service.

The usage-based pricing is meant to be more cost-effective for users than a do-it-yourself cloud deployment that requires them to license the MarkLogic software in traditional fashion, Krupa said. The vendor uses a similar pricing model in the MarkLogic Query Service, a query processing cloud option that was launched in July and can now be used in conjunction with the MarkLogic Data Hub Service.

The MarkLogic Data Hub Service will also be available on the Microsoft Azure cloud in the future, but Krupa wouldn't say when that is scheduled to happen.

Dig Deeper on Database management

Business Analytics
Content Management