Why We’re Building Datometry openDB

Struggling with your Oracle to PostgreSQL migration? Despite years of migrating, do you feel you only scratched the surface? The bad news is that what you’re doing is clearly not working. The good news is that help is on its way. 

In the next weeks, we’re kicking off the countdown to Datometry openDB, a revolutionary new approach to database migrations from Oracle to PostgreSQL. Sitting between your Oracle applications and PostgreSQL, openDB translates database queries in real-time. It makes an existing application written for Oracle “speak PostgreSQL” without the need to rewrite or modify it. 

Sounds too good to be true? We don’t think so. We’ve done it already — for Teradata. Now, we’re bringing the same capability to Oracle and PostgreSQL. Airlift empowers Oracle users to move from Oracle to PostgreSQL at unprecedented velocity and simplicity. 

In this article, I want to share what’s behind our initiative. Why are we building openDB for Oracle?  

Oracle had a good run. But their customers have complained for years about the cost structure and the price-performance ratio of Oracle. It was just a matter of time before it would reach a tipping point. Judging by the market dynamics, we’re seeing the first signs of this tipping point. Customers want to move off Oracle. 

Where do they want to move to? Enterprises are moving to PostgreSQL. At last, PostgreSQL is seeing traction across all industries. Because it’s open source, about a dozen commercial PostgreSQL editions are now offered by a wide spectrum of vendors. Customers can choose from an array of cost-effective alternatives to their once indispensable Oracle database. 

With over 30 years in the making, PostgreSQL has the maturity needed to take on the market leader. PostgreSQL is certainly not the last word in databases, but it’s a pretty good one to end a lengthy period of confusion.  

In the past, IT leaders were afraid that if they moved away from Oracle, they might make a wrong decision by choosing one database over another. That’s different this time. It seems everybody agrees there is only one alternative to Oracle left. And that alternative is PostgreSQL. 

Ask any Oracle customer and you will hear the same sentiment again and again: they feel trapped. They believe they are overpaying for Oracle, but the switching cost is even higher. Most of them will also tell you they tried every tool there is to migrate. AI is the latest entry in this space. At some point they need to ask themselves: why keep repeating what hasn’t worked and expect different results? 

What complicates things further is the fact that different groups within the organization have different ideas on how to solve the problem. Developers love the challenge of a migration because they want to write code, and a migration is the ultimate excuse to write code. Lots of code. Also, no other project has job-security written all over it like a database migration. 

Contrast this with IT leadership, which is much more outcome focused. Leaders look at it much more soberly. There is no free lunch, but PostgreSQL can cut their database bill in about half when accounting for infrastructure and maintenance. This plays well in an economic climate where everybody is looking for cost-savings. In large organizations, the move from Oracle to PostgreSQL may save tens of millions of dollars in annual fees very quickly. 

For those IT leaders who are looking for a solution, not a development process, Airlift is a cost-effective and efficient mechanism to unlock these savings. 

The realm of database migrations hasn’t seen much innovation in the past 40 years. By now, every database vendor has a toolkit that is supposed to make moving from their competitor’s platform to their own easy. The oldest dates back to the past millennium, Snowflake’s is perhaps the most recent one. 

Instead of belaboring technicalities, let’s look at it from an economist’s point of view. These tools have not addressed the problem effectively. If they had, Oracle’s stock wouldn’t be at an all-time high right now. And you probably wouldn’t be reading this article. 

The crux of the issue is that translating SQL cannot be done in isolation. Without the actual database running, many queries and directives are impenetrable for these tools. Therefore, database migrations turn into expensive consulting engagements that are not so much about converting applications but about rebuilding them. 

Because of the many manual steps involved, a conventional migration costs about 10x the annual database run rate. For example, if you pay $25,000 a year for your database, the switching cost will likely be about $250,000. This includes the cost of keeping Oracle alive while rebuilding applications, consulting fees, and less obvious positions like the costs of upgrading dependent systems that won’t work with PostgreSQL otherwise.  

The underlying calculations are surprisingly robust: a $2.5m Oracle estate will cost approximately $25m to migrate, and so on. The ROI on a migration is simply terrible. With Airlift, we’re changing this equation and putting the power back in the hands of the database owners. Migrate at 4x the speed, only 20% of the cost, and with none of the risk. 

The overall market dynamics and the realization that conventional migrations come with low ROI are what made us search for a better solution. openDB brings the concept of database virtualization to Oracle databases. 

openDB is for enterprises who don’t want legacy migrations but want results instead. By breaking Oracle’s vendor lock-in, Datometry levels the playing field so vendors can compete on the merit of their systems rather than hide behind switching cost. Let the games begin! 

About Mike Waas, CEO Datometry

Mike Waas founded Datometry with the vision of redefining enterprise data management. In the past, Mike held key engineering positions at Microsoft, Amazon, Greenplum, EMC, and Pivotal. He earned an M.S. in Computer Science from the University of Passau, Germany, and a Ph.D. in Computer Science from the University of Amsterdam, The Netherlands. Mike has co-authored over 35 peer-reviewed publications and has 20+ patents on data management to his name.