Shulou(Shulou.com)05/31 Report--

This article is to share with you about which six databases AWS has. The editor thinks it is very practical, so share it with you as a reference and follow the editor to have a look.

Which 6 databases does AWS have?

To be exact, I think it's seven databases:

* RDS

* RDS-Aurora

* Redshift

* DynamoDB

* Neptune

* Timestream

* QLDB

(some people think RDS and RDS-Aurora are the same database.)

These databases really have their own magical powers, capturing almost all the application scenarios related to the database.

Let's introduce them one by one:

RDS (Relational Database Service)

As its name implies, RDS is a "relational database". In fact, Amazon has transplanted several commonly used databases on the market and made a "cloud" version to customers, including: Oracle,MySQL, MS SQL, MariaDB, PostgreSQL.

The advantage of this is that customers have low education costs, low migration costs, use their own familiar databases, and enjoy high availability and high performance in the cloud.

These seem to be nothing, very "normal", not enough "insane", but the next few database products, AWS is about to open up.

RDS-Aurora

Although Aurora is a database product hanging under RDS, I think it is completely different. Cannot be compared with other RDS databases.

One of the core problems of the database is to solve the "high concurrency", including: high concurrency "read" and high concurrency "write". (for example, on an e-commerce platform, all inquiries about goods are read and orders are written. )

You might say that highly concurrent "reads" are not difficult to deal with! But how many more data parts will not do? For example, a piece of data is placed on 10 servers-- for reading, this is the case.

However, if there are many copies of data in the system, highly concurrent "writes" cannot be effectively synchronized to all replicas-so highly concurrent reads and writes are actually a contradictory combination.

Through the concept of "log is data", Aurora effectively separates "data engine" and "data storage", thus achieving an unprecedented high concurrent reading and writing function.

In traditional ordinary database services, or ordinary self-built database institutions, "writing" can only occur on one "master" data, and then the "master" synchronizes its own data to other copies. Unlike Aurora, "write" can occur in any availability zone. The architecture of Aurora uses three availability zones, each of which has two copies, that is, a total of six copies, all of which can be read and written. It greatly makes up for the bottleneck of high concurrency in traditional databases. Is this "insane"? How do you do this?!

I won't dwell on the details here. If you are interested, you can consult our architect.

Highly concurrent read and write is a typical scenario that occurs in OLTP (On-Line Transaction Processing online transaction processing). So what products does AWS propose for OLAP (On-Line Analytical Processing online Analysis process)?

Redshift

Unlike in the OLTP scenario, the database needs to support high concurrent read and write. In the OLAP scenario, the frequency of database read and write is very low, and the database needs a lot of aggregate computation: large amount of data and large amount of computation. (for example, at the end of the day, we need to analyze today's user behavior, and all user behavior data may be several TB. )

At this point, Redshift is needed to appear. Redshift is also a relational database, but it is fundamentally different from RDS in that it stores data not by "row" but by "column". Not only that, it also sorts the data by "column"! Basically, this is the database that was born for "aggregation". And the database aggregated by column is very convenient to compress, Redshift can deal with PB-level data!

You say this is fine: the traditional relational database, AWS; OLTP database; OLAP database. AWS doesn't think it's enough!

DynamoDB

The above are all relational databases. DynamoDB is a "non-relational database" provided by AWS, especially with "key-value storage" as the core. It can read big data / large files quickly and efficiently.

If you are in the game industry / medical industry / film and television entertainment industry with large files to read, then you are blessed. :)

Neptune

There are relational databases and non-relational databases! Isn't that enough? Of course, not enough for Amazon!

AWS also launched the "graph database". In this age of social applications (big data relevance), it is exhausting to take a simple interpersonal relationship and use a "relational database" without a graph database. Here are two simple examples:

1. We want to know how many times user An and user B are friends? If An and B are direct friends, just use the relational database Join once. But what if it's a 10-degree friend? You can't Join 10 times. The point is, if you don't know how many times you are friends, do you want to go down to Join indefinitely?

two。 Our logistics system reports that a truck from Beijing to Shanghai has been delayed, so how many users will be affected? Can one train affect 10 logistics links? Affect 50 cities? Affect 2000 users? It is estimated that we need to query all the data in the system before we can come to the conclusion that the cost is too high.

In the above scenario (that is, when a large number of Join appears constantly), the graph database is a sharp weapon.

There are already several databases, isn't it enough? AWS doesn't think it's enough. In 2018, it released two new databases at the same time!

I'm really going to kneel for this company.

Timestream

As the name implies, Timestream is a time-related database. Because many of the novelty searches of data are based on the time period. Timestream is aimed at this scenario.

QLDB

The full name of QLDB is Quantum Ledger Database. Its application scenario is also easy to understand, and the most suitable scenario is "account book". "account book" can not be changed, every record can not be changed, has been faithfully recorded, has been queried. QLDB came into being. Is it a bit of a "blockchain"? It's just that QLDB has a center.

Thank you for reading! This is the end of the article on "what are the six databases of AWS?". I hope the above content can be helpful to you, so that you can learn more knowledge. if you think the article is good, you can share it for more people to see!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.