Shulou(Shulou.com)06/01 Report--

What this article shares with you is what the data solutions for several types of relational databases are. The editor thinks it is very practical, so I share it with you. I hope you can get something after reading this article. Let's take a look at it with the editor.

Today, we will talk about the data solutions of several types of relational databases, which can be regarded as throwing a brick to attract jade, and we should also make some extensions to the technical direction in the near future, which can also be regarded as a summary of the early stage.

Oracle

At present, the mainstream version in the market should still be 11gR2. I remember a website did a survey many years ago. The proportion of 10g 11gR2 11g version is about 6:3. I think the proportion of 10g 11gR2 version should account for more than 90%, and the remaining share should be 12c. Nowadays, there are very few databases using 10g version, and there is no reason not to upgrade unless the business is stable enough.

Let's briefly talk about Oracle's plan.

From the perspective of disaster recovery, there is no doubt about Oracle Data Guard. From the basic function of 9i to 10g as a standard component, Far Sync of 11g Active Data Guard,Snapshot Standby,12c and more customization later, Data Guard has shown that backup database is not only a repository, but also can carry more business operations.

From the perspective of high availability, Oracle RAC is the leader in the current database high availability solutions. Because of a large number of production practices, this software is standard for many large enterprises. RAC is a typical share-everything architecture.

Of course, Oracle support sharding is also an indirect implementation of the original architecture, so in the late 12c directly made a larger-scale improvement, based on RAC can also do sharding.

Two other solutions are also implemented in 12c, 12c In-memory,CDB, which can do some resource integration and efficient data calculation on the database to a great extent, which can be regarded as a standard feature in 12c.

Judging from the characteristics of 12c, the improvement in the OLTP direction is very limited, and the force begins to be sent in the AP direction. Exadata is also one of them, the main direction is AP business, but because of the great advantages of hardware, some TP services are also very common.

So the architecture pattern of Oracle is share-everything, so no matter how large the database is, there is a limit. After all, the price of sharing is that there are some limitations in storage. Including AP business, can only be regarded as large and medium-sized data, there is still a certain distance from the massive data, after all, a hundred T database of TP business is a little worried, there is a kind of feeling that it affects the whole body.

MySQL

MySQL is characterized by the lightweight of open source, of course, light weight does not seem to be very strict now, now look at the branch Percona,MariaDB and other branches, there are many solutions.

MHA is a classic high-availability solution that is still popular today, and there are many customized versions in the community.

MySQL Proxy series

This version is currently the direction with the largest number of power points in the MySQL community. It can be seen that we have separated the architecture of the MySQL stand-alone bearer, using the power of the cluster as much as possible, rather than fighting alone.

Therefore, we will see a lot of middleware products on the market, whether commercial or open source, there are certain applicable scenarios.

MySQL Group Replication and InnoDB Cluster are official starting points. Now, Oracle is also sparing no effort to make great improvements, and many problems criticized in the past are now being repaired quickly, which to a large extent will not only put the original community products under certain pressure and risk, but also bring users closer to the official version. From my point of view, I am optimistic about the official plan, after the release of version 8.0, it will have a far-reaching impact on the current situation of MySQL.

In the direction of AP, there are two good solutions, such as the transparent switching of MySQL and NewSQL. TiDB is a good choice, which is now the main HTAP scheme. As a distributed solution, TiDB can be directly used as the Slave of MySQL.

And, of course, MariaDB ColumnStore, a later evolutionary version of the previous InfiniDB, is also worth a try.

PostgreSQL

PG is a little bee in the open source community and has produced a lot of good solutions.

I will briefly say two about this, one is for PG's cluster scheme GP, the MPP-based architecture pattern can carry more pressure and load.

The architecture of GP is slightly different from others.

There is also a popular database, CockroachDB, which is also worth paying attention to.

The above are the data solutions for several types of relational databases, and the editor believes that there are some knowledge points that we may see or use in our daily work. I hope you can learn more from this article. For more details, please follow the industry information channel.

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.