Zooming in on HTAP
Traditionally, online transaction processing (OLTP) systems, such as relational databases, and online analytical processing (OLAP) systems, such as data warehouses, are separated due to their differing performance needs and data models. HTAP eliminates the need to separate these workloads, enabling real-time analytics directly on live data. HTAP is transformative because it eliminates the need to copy data, which is time-consuming to maintain, prone to staleness, and can introduce inconsistencies.
Additionally, in traditional architectures, application teams typically only consider their operational system as the system of record, while data teams rely on the raw layer available in their data lake. With HTAP, all teams can align on a single, shared data store as the authoritative source.
However, this advanced capability—long sought after by data architects—has yet to become mainstream. As you can see in Figure 7.5, there are currently a handful...
