Thank you for everyone's feedback on this video. The intention of this video was to compare the UI and management tools for each of these, but the title and thumbnail was not representative of that. I updated the title and thumbnail to reflect it better. I do plan on diving into the actual databases to look at things like scalability, speed, and more. Thank you again for your feedback!

I have used all these, MS SQL Server and PostgreSQL extensively for decades, nearly all in large system / large data applications. MySQL has been used for small systems. Also used Oracle quite a bit, especially in the 90s. Review mostly focuses on the DB front end tool, rather than the DB capabilities, so a very misleading title / blurb to the video. There are multiple DB user tools, other than the standard ones supplied by the DB, including ones that work across multiple database types. The front end is a factor, but should NOT be the driving choice when looking at DB systems, unless you are a rank beginner, or have very simple DB needs. This review is NOT highlighting how to actually select the right database, as behind the toolset they offer very different capabilities. For example, Postgres has unparalleled support for time series, sharding and geospatial systems. PostGIS is the most powerful and widely used GIS DB extension / capability available on any DB type. PostgreSQL is ultra-extensible. The extension system is designed in from inception, and in addition to sharding, time series, and so on, also provides multiple database programming languages. In addition to pgplsql (the equivalent of MS SQL TransactSQL), you can program in Python, C/C++, and many others via suitable extensions. The list of extensions to support various capabilities likely runs in the 100s, and includes adding in different indexing and datatype options. MS SQL: I used this extensively as well. The big stand out here for me is Transact SQL. If you are doing complex, large data set manipulations, or are really locking down the database against misuse via the use of triggers, rules and so on (can do the same in Postgres), this can be a life saver. MySQL: their procedure / procedural support is very young, and nowhere near as robust as MS SQL or Postgres. MySSQL mostly came out of the embedded / small system / front end developer space, and has grown from there, where both MS SQL and Postgres have large system DB roots. (See ORM mention below) So, I would say, if you are a beginner / front end developer, who is not going to focus on the DB capability, and has a fairly standard requirements list. look at MariaDB (open source MySQL), but you may want to consider PostgreSQL depending on your target application environment. MS SQL is great, but not open source. ORM: The other use case for MySQL is if you are using the DB behind an ORM (object relational mapper), where the DB is just a data repository. Then it really doesn't make a huge difference which DB you select, as they are mostly black boxes. You should choose the DB that your ORM supports best, and is cheapest to own / operate. Obviously issues such as reliability and performance are important, and here Posgres and MS SQL have decades of experience to show. In our case, PostgreSQL was the only choice. Our application is heavily GIS (Geographical Information System) and time series focused, where we have to analyse 100s of thousands of records for a single transaction to provide the functionality we need. We also use pgplsql functions heavily for update and extract transactions as we have multiple front ends for the application. MS SQL was tested but the GIS component proved to just be too slow and not nearly as robust / functional as PostGIS. MySQL just didn't have the functionality to any useful degree. We could have used Oracle, but the costs were prohibitive....

There's very little comparison between the databases shown in this video. It would be better if you would show us how the databases excel one above the other in use cases. 80% of the video is comparison between MySQL Workbench vs SSMS vs PG Admin. These are actually tools to access these databases. You can access these databases with other tools like SQL Assistant. These are actually comparisons of tools not the actual database.

PostgreSQL does have a SQL extension it's called pg/plsql, it's very powerful, in fact, PostgreSQL is much more advance than MySQL only drawback is administration it's not as easy as MySQL and MSSQL

I learned SQL on Postgres and importing/exporting was frustrating if you got an error. I really enjoyed seeing the comparison of these platforms. Good stuff!

The quotation marks in PostgreSQL are only needed for establishing names to columns which could collide with system key words - for example a column called "name". And using the Schema name before the table is only needed when you have duplicate table names. So in both cases it's not a "weakness" compared to other databases, but simple features that can be used given a necessary scenario. And in terms of interface, all these applications run on the actual database, having multiple options for each engine, so again, it's hard to say the least which is the "better option", when it's a matter of personal preference.

MySQL is great but the best one is always gunna be the one your company uses... so Microsoft SQL server it is! 😂

Having to play with the 3. Nowadays I liking Postgress better. It was much better support to JSON and it provides a plethora of built-in functions for time series and others

Open source: You are not mentioning MariaDB, which is the open source fork of MySQL, and is becoming the default available in most Linux repositories, rather than MySQL which isn't true open source.

Well, you should compare the SQL language dialect differences, because some of the DB servers have the built-in programming language 'extensions' and they are really quite different. But the SQL language is quite standardized (by ISO , IEEE standard or something) and the difference between the SQL dialects is mostly the servers' built-in functions, like string or data values processing and some special language extensions which are using the specific features of each DB-server. For example, the MS SQL server is using the T-SQL (Transact-SQL) programming language extension, but the PostgreSQL is using the PL/pgSQL (Procedural Language/PostGres Structured Query Language) and the Oracle DB-servers are using the PL/SQL (Programming Language/SQL) extension. The PL-extensions are used for the server-based business logic programming, mostly, for writing stored procedures, user functions, triggers, packages, etc. So, in my opinion, you should compare the SQL language dialect differences, not the PL-extensions and not the IDEs (which is not a property of the SQL language itself).

Thanks @Alex for taking the liberty to explain the differences and similarities on different SQLs. Really informative. However, we don't choose what to use. It's the employer's choice.

My understanding, as a developer, is that as a data analyst you probably shouldn't really care about differences between different RDBMSs except query language and I hoped to see why I might be wrong in this video. But no, you just confusing RDBMS (Relational Database Management System such as MySQL, SQL Server or Postgres) and its client application (MySql Workbench, SSMS, PGAdmin4 respectively). So the only difference you mentioned between those RDBMSs is their query language. All other differences are really just some UI elements of their default clients. Anyone watching this, imo, should understand that distinction. You are not bound to single client app. There are many other clients, some of them even can work with several RDBMSs. And some of them are even can affect the query language! SSMS, for example, lets you write GO in the query editor as a delimiter between queries, but there is no GO statement in T-SQL, it's SSMS specific (their official command line client supports that too, I believe) delimiter, which makes your queries run as separate batches instead of one per editor tab. Some other clients give you more convenient features - Forge Studio for MySQL has quite good graphical query constructor. HeidiSQL is just very lightweight, can work with mysql, postgres and sql server, and probably something else, and it has freaking search bar in object explorer pane that even supports some regex magic (looking at you, SSMS, why with all Microsoft's resources there is no f-ing search bar right on top of object explorer??? Why search is buried in context menu and doesn't support regular expressions???). DBeaver is quite powerful and works with several different RDBMSs... That's just some clients that I remember and use or have used in past, there are surely way more of them. In fact, any application that connects to a database is a client for that database, just not every app lets you write queries yourself and shows results as boring tables. Some let you instead, I don't know, watch a video on your screen and read some comments below, and even write one yourself :)

I was just waiting you making a video about that! Thanks!

Thanks Alex! I used MySQL in my earlier years and totally agree with your points here :)

PostgreSQL is usually pronounced "post gress" and if you included the QL, you say " Q L" at the end, but since Postgres can also be a NoSQL DB (key pairs, JSON document storage, etc.), most people who use it regularly call it just Postgres..

Thank you Alex for the video. It was very timely as I had a bit of a confusion stint yesterday when I was practicing on a website.

Thank you Alex for keeping it direct and easy to understand. I worked for more than 20 years as a software developer since Turbo Pascal, Delphi, Visual Basic, Java, SQL, and after 8 years I'm getting updated to come back programming. Thank you for these videos.

Excellent comparison video - thank you!!! I used MongoDB (NoSQL) for a project because I wanted to be able to store ragged data formats for program preferences that wouldn't form neatly into a row/column table. I have since migrated to PostgreSQL since it has support for JSON and JSONB types which are even "indexable". Another big reason I went to Mongo was its ability to index spatial data (maps). Given Postgres's tight integration with mapping using PostGIS, it makes it a no-brainer for any map based database usage. BTW, I can't stand PG Admin, either when it was a compiled Windows/Mac program or now as a web based solution. It's slow and painful to use. I found a great multi-use SQL client, DBeaver, which in the free/community version supports many SQL versions and is lightning fast! Note: I have no relationship whatsoever to DBeaver - just found it, starting using it and became a huge fan!

I just love comparison videos. Thank you Mr. Alex.

Awesome! I was really looking for clear clarification like yours!

Overall I think this video is good exposure for people who are unfamiliar with database options, but its lacking on accuracy in some ways. For example you mention T-SQL and that its only available for SQL Server and while this is true, you failed to mention that MySQL and Postgres also have something similar. T-SQL is just the scripting language, in MySQL its just regular ANSI sql, but in Postgres theres PG/SQL which is heavily influenced by Oracles PL/SQL. Also you mention not being able to show T-SQL which was again not particularly accurate. The easiest way to do that would have been to show that you can declare variables and use them in queries. Again the same applies to Postgres. the other thing that I found could have been a bit better is how you mentioned the sys objects in SQL server. There are equivalents in both MySQL and Postgres as well. To me it seemed that you were not necessarily comparing databases, but rather comparing the administrative and development tools for each database type. I've worked with all three tools, and a few others and I find that MySQL workbench is the jankiest, but then again its an Oracle product so its to be expected.

I've used SQL Server at extremely large scale (taking in 1TB of data per day) and MySQL at very large scale. You get what you pay for. The performance, tools, configurability, SQL features, stability, etc are vastly better with SQL Server. It's not even close. I've only used Postgres in small instances but I would choose it over MySQL.

Ironically enough the MySQL in Steps book is the main one I am currently learning. It hasn't had any steps on learning how to import databases in the book. But that is ok. though. I do find it easier to learn. But I am going to go through the SQL Server one to understand the fundamentals too.

This is really helpful for me as I am studying and trying to get in the analytics field, knowing which to focus on is really useful. Thanks Alex.

I have 15 years experience with SQL and I think you explain it like there are big differences between them. You can very easily switch between them depending on the project you are working on. Unless you are the db Admin 🤣

Another great video! Your channel is truly amazing for someone trying to get into the world of data analysis (like myself)! Also extra points for the office reference in the data! ;)

Sorry, but those are the user interfaces that comes by default with the DBMs. Not a comparison by itself, which should start by how well they implement the relational algebra, and then how they implement the the desired properties about transactions, concurrency, indexing, consistency, security, permissions granularity, logging, recovery, escalation, mantainance, troubleshooting support, isolation levels and scopes, hinting and how much you'd rely on them, and so forth.

Converted an app from SqlServer to Postgres. One huge issue is that Postgres does not allow temporary variables (you'd have to use a function), while TSQL does. There's a way to do that sort of functionality in Postgres outside of a function, but it's not intuitive, and it's barely written up on the internet, so it took a while to get an optimal solution. That said, I do like Postgres a bit more - seems a bit more "pure". Postgres also seems to have a lot of datatypes. This might be relevant for some, though it's probably not an issue for the vast majority of apps. And yes, Postgres is free and SqlServer licensing is very complex and can get expensive (in the lower tier, you need server licenses as well as client licenses, and the client licenses are per device).

Thank you very much. Really clear and simple comparison done. Very informative.

got my first job as a DA! the work is incredibly hard though. first five tickets and it feels levels above me for entry level DA work. hope i can pick up their material and not be fired

Alex can you do more projects on sql n other data visualization tools like tableau, excel n python. What questions should we ask for analyzing before preparing vizzes.

Will focus on Posgres vs SQL Server in a quick post.Postrgres peformance is awful for complex sql with many joins, plus security is bad in most scenarios when you compare with SQL Server. This is from my experience with large tables. Things that take minutes and hours in postgres will take seconds in SQL Server. That's a fact. Also the default Postgres sql tool is not up to standard when you compare it to SSMS. Also if you are trying to locate a performance issue i have so many optons and tools in SQL Server but it's all very basic in postgres.

Quoted column names are also required in Oracle if the column names are mixed case. To avoid having to quote all required names in queries I always capitalise on import of data.

OpenSource does NOT mean it is free, just that you can see the code but might not be allowed to use it/modify it/execute it. This is a common and important misunderstanding.

Thanks Alex. I have more idea on mysql

Great Video! using MySQL just because of macOS

Super Informative video. Please make videos on how to write complex SQL queries. Thank you

Thanks for the video. I use SSMS with TSQL for work, but want to create something myself and don't want to pay. Although there are free options with Microsoft, I would like to not worry about scaling up. Think I might test out MySQL.

Also, just one more thing in today's NoSQL database trend: Postgres has really strong support for JSON / JSONB storage types, including indexiing. Use this quite extensively in my applications. It's not _just_ a vanilla SQL database, which is where MySQL probably scores well. MS SQL historically has also been SQL+ due to the emphasis historically on TransactSQL

I used SQL Server extensively from about 1997 to 2003. One thing I recall is they had was ability to view the entity relationship diagram which is important for design. Does MySQL server client have this as well? I really want to get into SQL Development again, on my mac.

SQL Server has advantage in diagram view, you can easily design the database at SSMS.

Alex , although you are 100% right but in poor countries the options as long they are free they are welcome :D big fan of your SQL mentorship from Egypt ^_^

You're a very effective communicator. Love your videos! Thanks, Alex.

I'm a MS-SQL guy but have a decent amount of experience with MySQL as well. I've always thought Management Studio was pretty awful when it came to general features from a UI perspective, when compared to something like dbForge Studio for MySQL. I WISH Microsoft would release comparable features (like having a table filter that supported multiple filters or even being able to colorize which DB you're looking at - small things). Ah well... some day, some day...

Hey Alex, can you make a video on importing and exporting data in each of these SQL?

Great video! Thanks! Well edited, you showed the interface and usage, the main differences, everything quickly and clearly, without wasting our time, and even gave an honest answer about which one you think is better instead of just saying "it depends"! Congratulations +1 subscribe

PostgreSQL is very powerful as far as spatial extension is concerned, it is really efficient and inter operable.

Not to sound harsh so I hope that you take this criticism to heart, but this video was more a comparison between SQL Clients (MySQL workbench, SSMS, PGAdmin) and not so much between the actual SQL languages. And not to mention that when you reached the PGAdmin part where you touched upon the dashboard tab, it became clear to me that you aren't actually that familiar with them. MySQL Workbench literally has something similar where you can see the server status (slow queries, queries per second, etc) which makes me think that you don't even know Workbench that well. And I am 100% sure that SSMS would have something similar as well. Also, there is a reason why postgres is a favourite amongst especially startup companies. Maybe do a bit more research on the topic next time. Again, I apologize if sound a bit too harsh. Just trying to share some feedback.

Can you do a video about data modeling and what a newcomer to data analysis should know about it, if anything?

Tns alex, your clips always get me out of the delimma

Great video! Thought I'd share my thoughts here on using MS SQL Sever on Mac, you probably might have heard this before -- it is possible to do so, using applications like Azure Data Studio or VS Code directly run on Mac, rather than on a VM. I personally like working in Postgres among the 3 :)

This is me right here... my college DB course textbook was in Oracle, the class project was in MySQL, and now my internship is SQL Server

Very useful video.. Thank you Alex. This is exactly what I was looking for!!!!

As a data analyst you don't need to write sql. Just connect to your data source with Power BI and use query editor. It's like a gui for sql.

Thank you so much, Alex. I have a question. I uploaded a data containing more than 1000 rows. I want to see all rows, but it always return 135 rows. How can it happen?

The first DB I use was .dat file and then db from dbase and goes to ParadoxDB, then I move to Interbase/Firebird and MySQL and totally move to SQL Server. For small business dbase n paradoxdb is the best out there. For medium Interbase/Firebird is a good move.

you're the best keep going dude

It depends on the project and how scalable the project will be in the nearest future.

MySQL for small to mid-size business with a regional focus. Postgress for National and International focus. Best for this type of relational large scale database. MS SQL for a National and International focus, plus the MS discount for mixing it with Azure and Office 365. Best for those all into using Microsoft only.

For common usage I agree. If you want most reasonable default setting on pages with low traffic MySQL will work fine without tweaking. BUT if you want o scale your project UP, Postgres on Linux OS is better because everything in it is configurable. Note: never trust anyone who tells you Master: Master geolocation separated replica works, and it is not advantage of MySQL, just a trap. Lost a lot of production data this way.

Tried using postgre and ran into problems with the group by clause. If I have * and an aggregate function in the select clause, I then have to state all the columns in the group by clause even if i'm only grouping by a single variable, let's say department. I do not encounter the same problem with MySQL which is so far my most preferred. Question, is it a must to mention all available columns in the group by clause even if grouping for only a single variable? Regardless of the SQL type?

I really don't know who the audience for this video is. The content in this is not what I expected, you should've included *for beginners* . You also should've mentioned that renting a server you'll have to choose Linux or windows and that to run MS SQL you'll have to rent a Wondows server which is more costly than a Linux server. Also writing APIs to access and consume the data plays a big part in which DB is best given your requirements and skillset. Sure same points are beyond the scope of a short video but the title is then misleading. Anyway my backend starter pack recommendation would be: - learn php - use mySql Easy to learn, setup, cheap and fast to implement and make changes.

Postgres delivers a level of features and performance that far outranks both MySQL and MSSQL. It comes out of the box with a bunch of features and optimizations such as native json support, gis support and pg/psql. I agree MySQL is a good "beginner friendly" DB, but for serious work I only typically use two databases; PostgreSQL (stand alone deployment) and SQLite (embedded).

Standard SQL is called “ANSI” Standard (DB2). PostgreSQL follows the “PL/pgSQL” SQL Dialect (similar to Oracles flavor - but Im reading DB2 now supports PL/1 too). Transact-SQL was created by a company called “Sybase.” Microsoft as is their pattern “partnered’ with Sybase and basically took their tech to make MSSQL-Server. To day MSSQL-Server is my favorite flavor of DB - but I’ve been in Oracle/PostgreSQL recently.

You are comparing database system that are different. The syntax is part of the tool but never should be the first choice when it comes to decide which database system we are about to choose. For example, mysql is normally not durable (unless they changed it); it means that you might loose some information in case of crash. So if you own a company that needs to be sure that no data is lost in case of crash and you choose mysql because it's easier to code/understand, you might get a lot of trouble when you server will crash and part of your data is gone. When it comes to maintenance, there is always a trade off between a system that has great support and a system that has less (paid vs free). Same comes with community. The bigger the community, the less pain you will have when looking for someone having a solution to your problems... These are important criteria too. For the rest, every database system has most in common in the language.

If you start using special functions in any SQL it becomes difficult maintaining cross SQL platform functionality. One of the reasons that TSQL becomes a PAIN. For enterprise level functionality Postgresql is very nice. Replication and pooling its hard to beat. Although if as you say you are just learning.... Mysql is a very good option or another that you didnt mention SQLITE... Super simple to install and GREAT for small applications..

I'd be interested to know a feature Postgres lacks, which SQL Server has... Postgres is what I suggest people replace Oracle with. MySQL is what I suggest people who want to use just SQL use, and I don't recommend anyone use SQL-Server

Been making a good living with T-SQL. I like MySQL because it's easy to pick up start working with any data or to use on your own projects.

I love love love google big query. It’s so clean and easy syntax. Sql Server is my second favorite

As a begginner, which one shoud l focus more or start with for thos two main sql MySql or Postgre?

Hello i found your channel yesterday & watched some of you videos they are really great !!! good job on !!!! all the information very helpful i just have one question based on your experience & this video seems to me that the way to go to learn SQL would be MYSQL i saw that you you have some good video's about Microsoft SQL tutorial. so my question is where could be a place to learn specifically MySQL ?? i looked for courseara but could not find some options for Mysql i found a lot for IBM SQL but not Mysql so any place you can recommend me to look for Mysql learning ? im very interested in Data analytics skills because that is the way im moving my professional grow. so thanks a lot & im watching today your tutorials for Tableau because I'm interested as well to learn this tool so thank you for your time & dedication for these videos that help a lot.

You seem to be relying on the frontend dashboard to evaluate these DB systems. For example, if u you use datagrip as your dashboard for Postgres your experience and evaluation would be different, ie better. Also, and more importantly, there are reasons to choose one of these DBs over the others based on concerns that have nothing to do with the frontend ui.

I've been using Postgres for almost a year and I never used quotation marks to SELECT. To be honest I didn't even know this was a thing

Are there major differences in the syntax of each of these?

Thanks for your honest review Alex...

Alex, before finding you out i learned MySQL and would like to do your portfolio projects in MySQL, can you please guide me a bit there?

Can someone pls guide me, which database is preferable for core banking domain between mssql and postgresql? apart from Oracle.

Useful video, but the audio sounds like it was recorded off Single Side Band. Excessive compression?

Even before coming across this video, my learning preferences for SQL was MS SQL then PostgreSQL then MySQL 😅

Thanks for sharing this. In my personal experience, PostgreSQL is the most fully featured and powerful Database out of the 3 you compared. PostgreSQL is the fastest-growing of all relational databases (a lot of innovation coming from PG community). There is also a plethora of extensions for many use cases and PG is being adopted as the core of many new cloud-native databases (neon, Yugabyte, Google's AlloyDb, CitusDB, AWS Aurora (Serverless as well), EnterpriseDB, and others). But I can't argue that the 3 of those are SUPER POWERFUL and loved relational databases. Keep sharing these great videos!

You probably should have differentiated better between the databases and the UIs. None of these databases are/have UIs; the UIs are separate third-party (except for Microsoft) applications. The pros and cons between databases are different than the pros and cons of the UIs.

Hi, mysql es down in the principal page. Any other page to download MySQL?

I have Microsoft SQL server and on pasting the code to create tables i get this error -A table can only have one timestamp column. Because table 'encounters' already has one, the column 'STOP' cannot be added. how to resolve this.

Did you just base your decision for which rdbms to get based on the frontend they ship? Pretty bad way of choosing a database system.

In terms of queries, Is it same for pgsql and MS server or else.

How do editors like PopSQL fit into this?

My biggest issue with postgre is importing data and the quotation marks. I had to create the table Schema with exactly same column before it will work.

viewer: Why we use MySQL ? Content-Creator : Because I love it . Really Bro!!

I'm getting post installation failed for database cluster while installing postgresql. Can anybody help please?

Do people compare all these with MS Access?

From my irrelevant and useless point of view, one of the most important thing is how fast you can query your data, how fast you can import your data, what can you do with your data, and how hard is to use your data. Taking about a program for how it looked is a little bit useless. To me, a it doesn't matter how pretty is the user interface for your database, the real important thing is the power and robustness behind the engine that keep my data. Not all databases are suited for all projects. For example, a lightweigth service that requieres a db and is running in a battery powered device (your phone) can't use something like postgresql or mysql embebed in the device, you use sqlite for that. In the other hand, if your software runs in a huge server located in a well powered datacenter, and you must process 500K transactions per second, perhaps is not a good idea to use sqlite. Before I saw your video, I was expecting to see some metrics and plots when you compared the engines in terms of inserts per second, reads per second, etc. I don't want to be rude in any way with your opinion, I just want to go to a point: You didn't compared the DB engines, you just compared the user interface. All of them are Database servers that programs can connect and work with bypassing the UI and go straigth via a remote client connection to do the work. UI are for humans and are no longer requiered once you application is up and running. This is my opinion and in no way must be take away as the truth or the only way of thinking. I would have chamged this video name to: which DB has the best user interface. I'm not a fan of oracle database, but is one of the most used db server where the data must be secured. Why? I don't know. you forgot to metion that here. have a nice day!

Dear Alex, Thank you for your videos. Is there any way i could ask you a question personally? I really need to know what you think about that. Stay blessed.

Talking about importing the files into server SQL server is the best we don't need to take care about the data it takes all the data but and no need to convert to csv if it's an excel file. comming to postgresql it's a pain we need to input all column names even after inputing all the columns there is no gaurantee that the file will import without errors we need to take care about the formatting of the csv such a large porcess And comming to MySQL the import is some what easy but when exporting a large data set it sucks and takes more time even after waiting for the import for me sometimes it doesn't import any data these were faced 8n MySQL workbench,for MySQL we can use dbeaver which is better than MySQL workbench and import time a lot faster

Hey Alex what about oracle SQL. I have about 200 questions on how to query Oracle database using SQL in a Windows operating system?

are you aware, that you're comparing not the databases themselves, but merely just some selected (by you) GUI tools (and there are much more gui tools out there to compare) to interact with the databases?

@AlexTheAnalyst Is there a version for mac users ?

Great review. Thanks.

in my opinion as PostgreSQL user, Microsoft SqlServer in another level With t-Sql which enable you to create awesome functions run on server , Which Done by programming languages in another database engines , MySQL is so popular because of powerful but because of easy reach and host , but in SqlServer send some variable by the name of stored procedures and let hem do the validation entries calculation and return you confirm or he rollback all changes as bulk of entries done by one procedures

What is the role of sql in data analysis?

Thank you Alex. You have helped me all the way through college, and now with my first interview in the real world. I appreciate all the help