Although I had a few problems with the original PR messages that invited me to meet with MarkLogic, the conversation with Gary Bloom, the company's CEO and president, was well worth the time.
Summary of our conversation
The following bullets are a quick summary to a complex and engaging conversation:
- The industry is experiencing several fundimental shifts in both the sources of data and how it is being used. The data is now coming from many types of end user focused devices, applications that combine the efforts of many systems that are housed all over the planet, neither enterprises nor end users will tolerate slow response times or failures, and older approaches that are based upon monolithic application and database design simply can't keep up.
- While it is true that things have changed in fundimental ways, older applications, systems and designs are not going away. They continue to support enterprise critical applications, but need help dealing with the tsunomi of data coming from everywhere.
- The state of the art in database architecture has shifted from a "shared nothing" design center to a "shared everything" center that can take advantage of local, virtual and cloud processing and data.
- Database design has moved from a "schema on write" to a "schema on read" approach to allow ad hoc requests regardless of how the data was originally delivered to the system. This requirement has emerged because enterprises no longer have the time or budget to invest heavily in reporting from multiple databases, developed to address the needs of a single application to address the requirements of a rapidly moving set of business requirements and government regulations.
- Enterprises have discovered, somewhat to their chagrin, that they must move at internet speed because no one is willing to wait for the more traditional static approaches to be executed.
- While enterprises are faced with new requirements, they cannot and will not uproot current applications and systems. They will instead "wrap" older technology and applications with new technology that will allow a huge acceleration in development while still providing the required levels of availability, performance and support.
A bit about the MarkLogic database
MarkLogic describes its database in the following way:
- MarkLogic is an operational and transactional enterprise NoSQL database that integrates data better, faster and with less cost.
- It is highly scalable and can grow as needed by adding more systems, virtual machines or cloud instances.
- It can pull data in from relational databases, mainframe databases, Hadoop filesystems and many other sources.
- The company claims it is easy to get data out of its database because the software supports a "flexible multi-model approach," allowing developers and end users to see the database in ways that support their needs rather than forcing them to use a view (schema) that might have fit the orginal application but doesn't fit today's use.
- MarkLogic's data supports Atomicity Consistency Isolation Durability (ACID) requirements, so transactions are reliable, safe and repeatable.
- Applications written for the MarkLogic computing environment can be run anywhere MarkLogic is supported. That includes, in the company's words, "the cloud, virtualized or on-premises. MarkLogic has years of proven success in the cloud and runs on AWS, Google Cloud, or Microsoft Azure."
Bloom and I go way back to the time that he was a member of Oracle's executive team and I was part of the IDC team of industry analysts. At the time of our first meeting, I was responsible for database software research for IDC's Unix and advanced operating environments team.
Bloom has always presented straightforward, pragmatic and logical arguments to support his statements. While I didn't (and still don't) always agree with everything he says, I do respect his opinion. It is clear that he has the customers' best interests in mind when speaking about his company's NoSQL database.
At first, I was ready to call BS on the messaging coming from this company's PR team. After a conversation with the good Mr. Bloom, it became clear that from their perspective, they were presenting the facts coming from their own customer engagements rather than just using hyperbole.
Enterprises are facing a crisis because of a time contraction. That is, they simply don't have the time to implement, redesign and re-implement functions just because a new government regulation or a new customer requirement has emerged. MarkLogic says moving to its database as a front end can easily address these requirements without imposing high levels of pain to developers or IT administrators.
I was impressed enough by our conversation to want to speak to one or more of MarkLogic's customers to learn more about what they were doing, what products they considered, why they selected MarkLogic, what benefits they've gotten through the use of MarkLogic's database and, I hope, any advise they'd like to offer everyone else.
My conversation with Bloom was well worth the time. You may also find a conversation with MarkLogic to be worth your time.
This article is published as part of the IDG Contributor Network. Want to Join?