Over a million developers have joined DZone.

Microservices and Data Virtualization

DZone 's Guide to

Microservices and Data Virtualization

Our microservices Zone Leader questions what the place could be for Data Virtualization in a microservice architecture.

· Microservices Zone ·
Free Resource

Everett Griffith had written recently an article about Microservices & Graphql as an API Gateway for microservices. Graphql really seduces me, because of the pattern it represents to me, and that I would call Data Virtualization.

Graphql proposes to aggregate datas from multi-sources, and to give access to data that is virtualized. No matter your data comes from a database, a rest service or a file, you access the same way to access your datas.

It gives then a direct access to datas to front-end developers, reducing the amount of time to create the api endpoint virtually to Zero. Ok I've not said that API is bad or not lean, but many of customers I've seen in my consultancy job can take up to one month just to conceive the API and to write the swagger. With this kind of solution, your Graphql call conception will not last more that the time to build a SQL request.

Everett Griffiths spoke also about the limitations of Graphql compared to API, as Authorizations and throttling policy that have to be developed.

I took a quick look to the "market", and it seems that not all solutions implement these mechanisms. I've found on a Forrester study that Redhat has a solution called "Jboss Data Virtualization", and I've found no evidences of it. You can limit some stuff like max-rows and some very technical and not Query related stuff.

I'll keep study these solutions, that are oftenly closed-source solutions, and I'll tell you more in the next weeks I guess.

Anyway, I do think that these solutions of Data Virtualization will maybe stay behind APIs for many companies, and could even be under microservices. The core advantage of this technology is to give a super kind of DAO to developers, fast to access, fast to develop, with easy caching. But maybe the future will tell me I'm wrong. And you, what do you think of?

data access layer ,graphql ,redhat jboss ,microservices ,data visualization

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}