10 problems with the Data Fabric architecture
A data fabric is a distributed data architecture with shared data assets, efficient data management and integration processes that are aggregated under one roof.
In a nutshell, data fabric aims to link any sort of data to everyone in an organization without having to physically provision servers or create large time-consuming IT integration projects instead of one data warehouse to rule them all. You have a virtualized view of your backend data sources using a data virtualization process accessible under one interface.
There is a lot of hype surrounding Data fabric and data mesh architectures, but it's not without its share of problems.
- Unauthorized access
- Having data access from one place creates a significant risk.
1. High upfront costs
2. High ongoing maintenance costs
3. Vendor Lock In
4. Underlying data sources must have pristine data quality and a robust master data managment (MDM) process
5. Dependant on data source, system solid uptime
6. Hire specialized SMEs needed to build write data virtualization coding layer
7. Security/Access concerns
8. Complexity in managing data fabric architecture
9. Gaps in data could be rendered in correctly
Say you have a customer in database one but is mislabeled or missing database two could give erroneous results upfront