5-year expectation of SQL Server

The beginning

In November 2005, Microsoft released another upgraded version of its database software, SQL Server 2005. This is the first time Microsoft has released a new version of database software in five years. Using the R & D capabilities of thousands of people, large-scale team development, modern software R & D management methods, it takes only five years to create a product, whether for Microsoft or any software company, are very rare. Let's briefly review the history of SQL Server over the past five years at this moment.

On July 6, 2000, Microsoft released SQL Server 2000 and announced that the next-generation database, codenamed Yukon, is under development.

In 2001, Microsoft released the technical details of Yukon, including the announcement of some exciting new features such as the next-generation storage engine WinFS, attracting the attention of many technical staff and the media.

In 2002, Microsoft released a. NET development tool, which was once mysterious. Net technology. Yukon, which is considered to be an important part of Microsoft's. NET strategy, was announced for the first time to extend its release.

On April 9, Microsoft again announced the postponement of the Yukon release date until the end of May, and then announced the postponement on April 9. In the same year, Microsoft released database components such as the report service and notification service, which will be integrated into Yukon and can also be installed on SQL Server 2000. This year, Microsoft released the first preview version of Yukon.

In 2004, Microsoft released Yukon a2. Microsoft confirmed that WinFS will not be released at the same time in Yukon and Longhorn, but will be released separately as a installable software in the future. This has left many FANS deeply disappointed with Yukon and Longhorn's high hopes.

In December 2004 and June, Microsoft released versions B1 and B2. October 30, 2005 Microsoft provides the MSDN subscriber with the download of SQL Server 2005 RTM. November 7, 2005 Microsoft officially released SQL Server 2005.

Here, it can be said that Microsoft has won a MileStone, and the development of Yukon is not over yet. For example, although SQL Server2005 provides database mirroring in Version B, it is removed from the official release, instead, it was provided as a separately installed component in the first half of 2006. Obviously, Microsoft believes that this function still needs to be tested.

Last Night stars last night wind

When Microsoft released SQL Server 2000 in 2000, I just started to write SQL Server 7.0 database applications using ASP/VC, I still clearly remember that the first task my company gave me was to save images to the database in Web mode.

In 2001, in order to write a project's second-phase plan, I began to gain an in-depth understanding of SQL Server 2000. However, compared with SQL Server 2000, SQL Server 7.0 has many obvious differences, so that for a long time I have been worried about how to persuade very rational customers. In the end, the customer agreed to upgrade to SQL Server 2000 a year later. After we proposed some long-term plans to use SQL Server 2000 for development, the system was upgraded.

In 2003, due to the needs of the project, my development team and I created a good data presentation and analysis product, strong historical data processing, analysis, and display functions are highlights. At this year's Tech-ED conference, I saw a presentation of Microsoft's new report service. Objectively speaking, at that time, the report service was still very rough. In terms of the friendliness of the interface and the practicability of the analysis function, there were still some gaps with our products. However, a good idea of the report service is to allow business personnel to design their own reports. It clearly divides the layers of report development, management, and usage, although it also requires more improvements. However, this is something that our technical team with limited resources and manpower cannot afford for a long time. After all, data presentation is only a part of the products we provide, even though it helped our products win praise from our customers that year. It can be said that focusing only on one aspect and making continuous improvements is a major factor in the end for the report service to outperform our products.

Many developers I have encountered are very concerned about how to implement the problem. Our product also includes a complete and flexible data processing and conversion, including the metadata definition section, the data warehouse section of SQL Server 2005 also provides the ability to conveniently extract, convert, and load data from multiple data sources.

This year, due to recommendations from friends, I wrote some chapters in an SQL Server book for a publishing house in my spare time. I remember that when I promised my friend, I was confident and started to write it, but I found that many places were still vague. I had to read a lot of documents and materials, even read SQL Server books from start to end, and finally handed in the books at the specified time. This experience has helped me find that there are still many subtle points in SQL Server 2000, which will be very helpful for the development and management of SQL Server.

In 2004, we provided a telecom company with a technical analysis on customer consumption behaviors. Although all sample data extraction and conversion are completed in SQL Server 2000, SQL Server 2000 only provides two data mining models when selecting a Data Mining Model, I can only use some free data mining tools on the market, as well as a considerable number of manual calculations to complete the final data discovery work. At that time, Yukon provided a wide range of data mining algorithms.

Looking back at the five years of close contact with SQL Server, from basic SQL Server application development skills to the theoretical knowledge of SQL Server, from simple use to in-depth use, I feel that SQL Server's existing functions are not enough. From the perspective of learning knowledge, it takes too long. To some extent, this is also related to the lack of in-depth analysis, research, and introduction of SQL Server Data in China. Almost all books stacked in the bookstore can be summarized in three words: full and shallow. In the summer of 2005, I finally read a good SQL Server 2000 Development book in the bookstore. It was written by Hu Jian, the first master of the Chinese SQL Server Technology Forum. However, it is too late for me to sigh.

In addition, for SQL Server development, almost the vast majority of Chinese developers think it is very simple. It is easy to develop SQL Server. However, most of the projects I encountered problems with SQL Server, and some even affected the overall satisfaction of the project to the customer.

Now, SQL Server 2005 is released, which is far more complex and powerful than SQL Server 2000. How much time does it take for you and me to deeply use and master this system, when can I see a comprehensive and in-depth SQL Server 2005 book in China?

But in the dark

From SQL Server 2000 to SQL Server 2005, not only does the database system have higher performance and processing capabilities, but the new version of the system also brings many new features that never appear in the old version. These new features are also consistent with the development direction of modern databases. For example, support for XML and support for embedding advanced languages in SQL languages. In the enterprise environment, more emphasis is placed on security, high availability, and integrated management tools.

The following is a simple list of new features and important improvements in SQL Server 2005.

1. Support for XML data types and XQuery queries
A few years ago, one of my friends was the technical support of the world's only pure XML database vendor. Driven by him, I was also very concerned about the XML database for a while. However, it turns out that although the support for XML is very important, the database must support relational data. SQL Server 2005 provides the most perfect solution in this regard.

2. Write database objects using CLR
Other large database systems, including Oracle and DB2, have now begun to support writing database objects directly using advanced programming languages outside the SQL language. There are two main reasons: the SQL language itself is still relatively weak, and advanced languages can be used to achieve more powerful functions. In addition, advanced programming languages such as Java and C # already have a very large development team, the direct use of advanced languages to develop database objects greatly reduces the development difficulty and program maintenance difficulty for these program developers.

3. UDT
In SQL Server 2005, user-defined data types are supported. Just as object-oriented databases are now regarded as the direction of next-generation databases, how to use UDT depends on the imagination of developers.

4. Service Broker
It provides a stable, reliable, and efficient Asynchronous Message Queuing queue.

5. ADO. NET 2.0
ADO. NET also has many new improvements, such as support for distributed transactions among multiple heterogeneous databases, new connection Provider Program models, and support for asynchronous programming.

6. T-SQL Enhancement
In addition to some more powerful SQL functions, some new content including support for exception handling is also added.

7. Database Mirroring)
A hot standby solution designed for key enterprise applications that focus on availability can greatly improve the high availability of SQL Server and switch the load to the standby system within seconds of Database Failure.

8. Full-text index Enhancement
Google's success shows that in the age of rapid increase in information, it is imperative for enterprises and individuals to quickly and efficiently find the data they need. Full-text indexing provides support for storing a large number of documents in the database and quickly searching for these documents. SQL Server 2005 significantly improves the storage and search performance of full-text indexes compared with the old version.

9. Integration and enhancement of management, monitoring and development tools
Integrate management, monitoring, and development tools to facilitate the DBA of an enterprise.

10. BI Enhancement and Improvement
If SQL Server 2005 is used for the customer behavior analysis I mentioned earlier, I believe it can be easily implemented. In terms of data mining, in addition to nine classic data mining models provided by the system, it also supports custom model algorithms.

11. Storage engine improvement
In the core storage engine of the database, SQL Server 2005 has made a lot of improvements, so the overall performance, stability and reliability are significantly enhanced than SQL Server 2000.

Due to space limitations, I cannot list all SQL Server 2005 improvements here. However, we can see that SQL Server 2005 is indeed a commercial database system suitable for key applications in the enterprise environment. I believe that a powerful and stable basic platform will surely bring tangible benefits to the informatization of industrial users, at the same time, it will provide more confidence and capabilities to software development and system integrators.