Managed databases, analytics, and costs. In this lesson, you will learn how to choose the right AWS database offerings, how to choose the right AWS analytics offerings, and how to monitor your AWS costs. At this point you might be asking two questions: First, what is a database? And second, why don't we just store all of our information on a service like S3, then query S3 using Athena. A database is an organized store of information. We still use databases today because we can perform queries, calculations, and other operations inside a database faster than on other storage services, especially across massive amounts of data. Your next question might be why does AWS offer its own database services, when I can simply run a database on an EC2 instance? The answer is that's true. You could absolutely buy your own database license and install that database on your EC2 instances. However in doing so, you'd be responsible for managing that database. You need to configure your database properly, apply patches, install upgrades, and manage your database backups. What AWS offers with its database services are managed databases. With a managed database, AWS takes over all the undifferentiated heavy lifting such as configuration, patching, upgrading, and automating backups with AWS backup, leaving you to focus on using the database to meet your organization's business objectives. Databases come in two main languages, SQL (or relational databases) and NoSQL (or non relational databases). SQL databases look like an Excel spreadsheet with many, many interlinked tabs like this visualization here. NoSQL databases look like a pile of Word documents, but each document might share some common fields such as the title, the author, or the date. The majority of databases in use today are SQL or relational databases. SQL stands for structured query language and SQL databases have been in use since the 1970s. Amazon offers two services for manage SQL databases: RDS and Aurora. Amazon Relational Database Service, abbreviated RDS is a SQL database service. You choose a DB instance type which is similar to an EC2 instance type. For example, you can choose small, medium, large. And then you choose the database engine that you'll run on top of that instance type. A database engine is a different implementation of a SQL database. RDS supports this list of SQL database engines. Database engines are either proprietary or open source. To use proprietary engines, you must have a license. To obtain that license, you can either buy it from the AWS Marketplace or you can bring a license that you bought elsewhere, for example, from the company that made the database. To use an open source engine, you need not buy a license, you just simply select it and start using it on AWS. From this list, the middle two engines, Oracle and Microsoft SQL Server are proprietary databases. MySQL, MariaDB, and PostgreSQL are open source database engines. The last engine is Amazon Aurora, which is proprietary and another AWS service. Amazon Aurora, which is one of the engines that runs on Amazon RDS, is a database engine performance- tuned specifically for AWS. It achieves up to five times the speed of a standard MySQL engine at 1/10 the cost of a commercial database. At time of recording, Aurora only works with two SQL implementations, MySQL and PostgreSQL. One Amazon Aurora customer is Netflix. Netflix streams 15,000 titles to over 200 million subscribers across more than 190 countries. More than 62% of American households subscribe to a Netflix account. When working with a MySQL database with billions of rows, Aurora helped Netflix reduce query time from 32 minutes to 3 minutes to find the correct data in the correct row. What if you wanted to migrate a database from on-premises to RDS? Alternatively, what if you wanted to migrate from another database engine to one supported by Aurora such as MySQL or PostgreSQL? Then you would use AWS Database Migration Service, abbreviated DMS, to perform these migrations while keeping your database available for users during the entire process of the migration. And speaking of DMS, there's also closely related AWS Server Migration Service, abbreviated SMS, that migrates on premises applications and server workloads to EC2.
