Understanding Data Virtualization Software and Its Key Insights

·Jan 31, 2023 11:21 AM

analysts have materials about data virtualization software. they talk a lot to the ppl who buy the software, so they can be a good resource for understanding what's important when evaluating. they also have a pretty comprehensive list of technologies in the space. There's also api virtualization, like the framework apollo, which comes at some of the same problems from a different angle. to your questions about virtual graphs, for stardog, yes, entities can be split across sources.

🙏2

14 comments

· Sorted by Oldest

François S.
·
so it is either virtualization of rdbs or virtualization of APIs but not both?
Michael G.
·
w/in a single application? both approaches are common in the enterprise these days
François S.
·
YEs, within the same KG
Michael G.
·
from a data source type perspective, absolutely. the data about a Customer, or whatever else is important to the business, is stored in every technology ever invented in a big enterprise
François S.
·
In most enterprise information systems, data is coming from multiple sources available through different means. for example Atlassian products are accessed through an API, and other systems through a rdb
Michael G.
·
the hard part about an API like Jira's REST API is that they were never meant to be used as what effectively is a table scan, from the perspective of a virtualization engine
Michael G.
·
if you require data from multiple different rest apis, you have to be realistic about performance expectations
François S.
·
got it, the system would have to maintain indexes of some sort
Michael G.
·
which gets down to the basic con of data virtualization. the data is somewhere else. that's unavoidable latency.
Michael G.
·
which then comes down to an engineering dance balancing data recency with latency. ie most ppl throw some sort of caching mechanism at the problem.
François S.
·
got it
François S.
·
thanks Mike!
Michael G.
·
no problem. see you in a couple months 🙂
François S.
·
🙂 🏙️