Return of DTCC-highly available and scalable database architecture

I mentioned that I will attend the 2010 Database Technology Conference. Today I will post the slides used in this conference, which may be useful to everyone. Let's take a look at some of the situations of the next Conference. A total of 29 lectures were arranged during the 2-day conference, covering Oracle, MySQL, DB2, SQLServer, Sybase, DM (domestic database) and other databases

I mentioned that I will attend the 2010 Database Technology Conference. Today I will post the slides used in this conference, which may be useful to everyone. Let's take a look at some of the situations of the next Conference. A total of 29 lectures were arranged during the 2-day conference, covering Oracle, MySQL, DB2, SQL Server, Sybase, DM (domestic database) and other databases

I mentioned that I will attend the 2010 Database Technology Conference. Today I will post the slides used in this conference, which may be useful to everyone.

Let's take a look at some of the situations of the next Conference. A total of 29 lectures were arranged during the 2-day conference, covering Oracle, MySQL, DB2, SQL Server, Sybase, DM (domestic database) and other databases, with a large number of lectures, rich topics, rare real numbers, haha. The lineup of speakers is also very large. Nine of the 14 domestic ACE/ACE ctor s were present.

You can download the PPT for each topic. Here I will introduce the topic "highly available and scalable database architecture:

In the topicHigh AvailabilityAndScalableWe have discussed the following two aspects:

High Availability

• High Availability of software/hardware (hot/cold standby)
• High data availability (sharing, same/asynchronous replication)

In addition to redundancy, the high availability of hardware alone is not much to talk about.
The high availability of data based on shared devices is only a rough introduction of possible solutions. Because the implementation of various solutions is expensive, it is more suitable for Oracle and DB2, so it is not discussed in depth.
Therefore, this section focuses on the use of MySQL Replication technology and applications to achieve high availability with the Share Nothing method.

Scalable

• Scale Up)
  • Hardware resizing (Increase the number of CPUs, memory capacity, and disk capacity ...)
  • Hardware upgrades (replacing higher-end hosts, higher-end storage devices, higher-end CPUs, and faster disks ...)
• Scale Out)
  • Data copy and distribution (multiple reads in one write, read/write splitting ...)
  • Vertical sharding and horizontal sharding)
  • Cache and Search (application update Cache, database update Cache, use Search full-text Search ...)

For scalability, Scale-Up is basically a competition for the expansion capability of each vendor's own single device. We do not have much capability intervention, So I simply analyzed it.
For Scale Out, I think it must be the most important issue for everyone. The Sharding in Scale Out is certainly not the first time you have heard of it. After all, it is nothing new.
Here I will focus on how to select the appropriate Sharding method in the Sharding process and how to solve the data merging problem after Sharding (which is not actually solved, shard ...), And how to use external database resources (Cache, Search) to solve the scalability problem of the data layer.

In fact, the architecture itself is benevolent, wise, and wise. There is no omnipotent architecture, and there is no architecture that can be applied for a long time. The architecture is closely related to the business scenario. If we leave the actual business scenario to talk about the architecture, it can be said that we are talking about it on paper. What if we leave the architecture to only pursue fast business implementation? Well, you have to pay it back sooner or later.

Note: I am not an architect, and I am very touched by taking so much valuable time to listen to my architecture...

High-availability and scalable database architecture solution

View more presentations from Sky Jian.

Original article address: DTCC return-high-availability scalable database architecture discussion, thanks to the original author for sharing.