Great educational post by Robert Hodges about what is a data fabric, the architecture of data fabrics and underlying high level implementation details:
It must hide failures, maintenance, and schema upgrades on
individual DBMS hosts. It must permit data to distribute across
geographic regions. It must deliver and accept transactions from
NoSQL, data warehouses, and commercial DBMS in real time. It must
allow smooth technology upgrade and replacement. Finally, it must
look as much like a single DBMS server to applications as possible.
I’ve learned a couple of things from it:
- the expectations and requirements, plus possible approaches for designing data-as-a-service systems
- the challenges, complexity and trade-offs involved in scaling relational databases.
Original title and link: What Is a Data Fabric Architecture? ( ©myNoSQL)