What are the benefits of NoSQL?

Question

NoSQL

I can not understand. Well, if you want, then serialize the data and store in MySQL the same as in Mongo DB...

Answer 1

Benefits and Drawbacks of Relational PostgreSQL vs MongoDB
https://youtu.be/SNzOZKvFZ68
In short:
Benefits of PostgreSQL - full transactions, works well with complex queries.
Disadvantages of PostgreSQL - does not work well when scaling out on a large cluster.
Benefits of MongoDB - well behaved when scaling out, fast on simple queries.
Disadvantages of MongoDB - poorly maintains data integrity in a cluster, provides data correctness not at any point in time on any server, poorly queries on different types of documents (different tables)
Previously, MongoDB had another advantage in schemaless, but with the advent of JSONB, PostgreSQL has this Monga no longer has the advantage.
Use Monga for non-important data of secondary importance, where scaling by cluster is needed.
Apply PostgreSQL on important data (for example, financial).

Answer 2

In short, NoSQL is everything that is not SQL . Something that does not fit the notation of the relational model.
A NoSQL solution can be beneficial when there is no need to follow a specific schema (tables, relationships, data types, etc.). We don't think about preliminary adjustment of the structure, normalization of relations; data can be easily changed, migrated, etc. But instead, there are problems with ACID and other things that are easier to observe in SQL solutions.

Answer 3

They have no advantage, they are trite for other purposes

Answer 4

well, if you really want to, then serialize and store)
tools of the same class are interesting because they solve the same problem. And if you have always used one tool, then it may not be clear why all the others are needed.
But, you know, from the fact that you most often use one specific one in a set of screwdrivers, all the others do not cease to be needed.
Mongodb is used for high load projects.

Answer 5

In fact, noSQL can be very different, sharpened for completely different tasks, which are very poorly stretched on regular SQL with its transactions, logging, indexing, and query system.
I will give a few examples where noSQL wins at times. And, in all tasks it is possible to be perverted and on normal SQL.
Imagine the task of walking through graphs to analyze social relationships. In SQL, we will get a couple of dozen tables, or take a graph database.
The task of collecting and aggregating (ordinary counters) - you can also take SQL, but usually there is no need for transactions and the associated overhead, so we take a tarantula, radish or kaffaka.
The task of full-text search in documents - solving it in SQL is not a very pleasant task, let's take solr / elastic / sphinx for example.
The task of storing objects is also difficult to solve in SQL, especially when the structure of the objects themselves is greatly expanded. Let's take Monga.
The task of storing time series logs with aggregation is just seams with SQL! Take influxdb or clickhouse.
At the same time, SQL very often has a serious drawback in scaling, which is easily solved in noSQL.
Oh, yes, very often they put both SQL and noSQL side by side, and the data is scattered into different databases for specific tasks.