Reasons Why HTAP Is a Pie-in-the-Sky Vision: 

  1. Transactional databases accommodate a small amount of data. 
  2. Transactional databases require minimal queue time and freeze data. 
  3. Transactional databases accommodate data with a high probability of access. 
  4. Analytical databases support large amounts of data.
  5. Analytical databases support large analyses. 

We partnered with American computer scientist Bill Inmon. Known as the “Father of the Data Warehouse,” he has much to say about HTAP and why it’s nothing more than a pie-in-the-sky vision of sorts. 

Table of Contents 

  1. What Is HTAP? 
  2. The Problem With HTAP   
  3. Integrate Your Data Warehouse with Integrate.io   

Sign up for a free 14-day trial, then schedule a trial set-up meeting with us. That way, you’ll know exactly what to expect. 

What Is HTAP? 

The dreamers of technology have long tried to make a single technology that suits all needs.  

IBM tried with DB2 and failed miserably. Big Data made a halfhearted attempt. Others have sought to have a single technology that serves all purposes. Now comes along HTAP.  

The vision — if you can call it that — of HTAP is to have a transactions processing database and an analytical database all in the same place. 

Related Reading: The Importance and Benefits of a Data Pipeline 

The Problem With HTAP 

The problem is that the developers of HTAP are trying to reverse some very basic laws of physics. The developers of HTAP are trying to say that down is up and up is down. There are some fundamental architectural reasons why down is still down and up is still up, regardless of what developers say.

Consider the following differences in characteristics between transaction and analytical databases. 

Transaction Databases 

To build and maintain a transaction-based system, it is necessary to: 

  • Accommodate a relatively small amount of data. Response time suffers if you try to accommodate too many data transactions. 
  • Require minimal queue time for the waiting of transactions in the processing queue. When you mix large analytical transactions with short operational transactions, at some point, the small operational transactions must wait on the large transactions to complete processing. And transaction queue time is greatly impacted by this interruption. 
  • Freeze Data. Operational transactions that can update the system have to lock data in place to ensure processing integrity. The freezing of data has a very deleterious effect on the performance of analytical transactions. 
  • Accommodate data that has a high probability of access. 
  • Support online access and update of data. 

Analytical databases have an entirely different set of characteristics. 

Related Reading: How To Choose the Right Data Integration Strategy for Your Use Case 

Analytical Databases 

Online analytical databases are characterized as having to: 

  • Support very large amounts of data. This is necessary to accommodate the analytical needs of the corporation. 
  • Support large analyses. The end-user analyst operates in an experimental mode in many cases. The end user analyst needs to be able to access an unfettered amount of data. 
  • Handle the probability of access to analytical data. 

Related Reading: Consolidate Your Data on AlloyDB With Integrate.io in Minutes 

Integrate Your Data Warehouse With Integrate.io 

Transform your data warehouse into a data platform by integrating all of your data into the cloud. Integrate.io offers a no-code data pipeline platform featuring data observability and data warehouse insights. Your data warehouse is powered using ETL and Reverse ETL to design data pipelines quickly, ELT and CDC for fast real-time data replication, and API Management to instantly generate APIs. 

Schedule a data warehouse platform demo and get started today! You’ll gain access to 100+ data sources and destinations, so you can try it out for yourself. 

Bill Inmon, the “Father of the Data Warehouse,” has authored 65 books and was named by Computerworld as one of the ten most influential people in the history of computing. Bill’s company – Forest Rim Technology – is a Castle Rock, Colorado company. Bill Inmon and Forest Rim Technology provide a service to companies in helping companies hear the voice of their customer. See more at www.forestrimtech.com