Hello, NoSQL becomes more and more popular (and actually is easier to manage than SQL database). So is there a possibility that MongoDb be available here as MySQL alternative?
Thanks 
You can try using Mongo DB atlas:
I don’t think they will add it because it’s lamp package and most people don’t use mongo with php.
Almost all PHP software out there still relies on some kind of SQL database to work, and almost all of them prefer or at least support MySQL.
I agree that NoSQL databases are becoming more popular, but NoSQL databases are an extremely broad category of services. I think a key-value store like Redis or Memcached makes more sense, followed by a search engine like Elasticsearch or Solr. Those are things, say, a WordPress site can make use of to enhance functionality.
MongoDB more or less replaces MySQL, but with worse tooling, worse script support, little schema, no relations and poor multi tenancy. So consider the possibility of us offering MongoDB to be virtually absent.
1 Like
So pity, as popularity of SQL databases is just an artifact of legacy. With NoSQL you get flexibility. And i am sure, should NoSQL option be available, it will outpass SQL.
They do not use it because it is not available, dear Pikachu
I strongly disagree with you on that statement.
SQL databases are old, but definitely not outdated. Most business data is inherently relational. Every web application has relations between Products, Categories, Orders, Customers, Payments, Addresses and so on. You generally want strong schemas, relational consistency, ACID compliance and so on to keep your data consistent.
Sure, schemaless databases are “flexible”, but that just means you’re going to have to protect the data consistency yourself. And deal with the consequences if (or when) you fail.
It’s no surprise that new NoSQL like databases which function like relational databases are gaining traction. CockroachDB for example is compatible with PostgreSQL and is much easier to scale, but is lacking in both features and performance.
Also, it’s not like we can offer a NoSQL database and have “NoSQL” covered. Some want a key-value store (e.g. Redis), some want a document store (e.g. MongoDB), some want a column store (e.g. Cassandra) and some want something else entirely. You can’t offer one type of NoSQL database and satisfy everyone.
Sure, the Facebooks and the Netflixes of the world have NoSQL databases. But they also have thousands of engineers, and datacenters all across the globe. Most websites don’t have such needs. Just because Netflix does it doesn’t mean you should too.
3 Likes
Actually, in the TOP 10 of database servers popularity there are 7 SQL database servers, 2 NoSQL, and… Microsoft Access (that supports SQL).
1 Like
That’s right, you get more code side responsibility for data consistency and that’s it. Your code starts to manage data relations, but not a scheme, which is flexible, scalable and easier to rearrange should there be a need.
It is just a different philosophy.
MongoDb (NoSQL Document) is number 5 being 12 years old, while number 1 Oracle Db (SQL Relational) is 49 years old.
NoSQL is climbing
Not always because a thing is old also means outdated.
You say it likes the data consistency management comes automatically. It doesn’t. In my experience, it’s really hard to do it well, and be able to cover every edge case with every possible failure scenario you get in a real world environment.
Schemaless data does not exist. In your code, you’ll still expect that, say, a Customer has certain attributes like a name, street, house number, city, etc. You can’t just have some blob of randomly named attributes and have an application make sense of it.
And in MySQL, I can just rearrange my table with a single ALTER TABLE statement and MySQL will update all the rows for me. MongoDB doesn’t do that for you.
That’s not climbing, that’s stagnating.
It’s been 12 years. 12 years is a long time for many new applications to be developed with the technology. If it really was obviously better, it would be at number 1 already.
If you look at ones which are actually climbing, you can look at number 16, which was at place 23 last year. Which is Azure SQL Database. Which, you guessed it, is a relational database.
MongoDB didn’t climb at all. But MariaDB, which is a fork of MySQL, did rise one place.
In the trends page, you’ll see that PostgreSQL is rising slightly faster than MongoDB.
So I see no evidence whatsoever that MongoDB or other document databases are overtaking SQL databases.
1 Like
This topic was automatically closed 15 days after the last reply. New replies are no longer allowed.