data base paradigms part2
Welcome back let’s continue. We have seen the
relational database. But what if instead of modeling
the data as schema we use it a as data?😕
5-grpah:
Our data now will be represented as node
and the relation as edges. Example: neo4j and Dgraph
In SQL to make a many-to-many relation we
use an intermediate table to join. In graph it’s easier! We just create edges
and link it to other records.
This
will lead to greater performance in compare to SQL especially for great number
of joins or data. It’s very efficient for recommendation system or when example
like Facebook friends with comments and posts …a very example here is for recommendation
engine powered by airbnb.
6-Search:
Meili search or angolia. Pretty sure you
haven’t heard of. In a nutshell as if the title explain it’s used for
searching. We provide a text and we get the possible result. It’s a bit like
the document oriented db. The difference is that under the hood is that is created
index for searchable term similar to the book’s index. This type is very fast but
it’s expensive when scaling.
7-the multi-model:
The most exiting one, The best example
here is Fauna DB used by twitter. In front end when consuming API you can
describe how you want your data using Graphql. Under the hood Fauna works to
take advantages of graph, relational and document. This all makes it a very string one and a
great candidate to try.
It’s great for almost everything.
Well there are other types: Time-series
and data warehouse… why not take a look on it?
Understanding the paradigms of a database
can save you project or even your job other day.😉
Hope you like please give me feed back.😍
No comments