XML is a flexible language. In which the user can mark their own tag (tag) to define their data type. In these tags, the user can define the principle or the name of the attribute, so that if the XML file is obtained, the data can be easily read by another user who has no knowledge of the system.
When different users are exchanging data, as long as the XML standard is followed, each application can write an XML interpreter for the data type it uses, easily converting the data in the XML file into its own data type, thus ensuring that the data is communicated quickly between different applications.
But as a database for storing data, previous relational tables are not well adapted to this new requirement. Users still need to spend a lot of time translating XML data into relational types and then entering the database, or writing complex queries as needed to read the data. This approach both adds difficulty and reduces efficiency. The original XML file will be parsed into the tables, which is a fairly resource-intensive process for a system that is complex or has a large amount of data.
The DB2 9 database has a significant innovation relative to other database management systems. In fact, the DB2 9 support for XML is not only embodied in XML-type data, but some problems that are difficult to solve with relational models can also be solved using XML models. The advantages of DB2 9 for XML support compared to classical relational data models are not negligible.
Compared with other mainstream database management systems, DB2 has its own advantages and disadvantages. The performance of the DB2 on the processing datawarehouse system should be excellent, and DB2 is perfect for the optimizer, which can be effectively rewritten as the optimal statement for most complex queries and allocate a reasonable execution plan.
There are some deficiencies in the mechanism of DB2 closure. This is related to the design framework of the DB2, the use of memory locks in the promotion of efficiency, but also on the system optimization requirements mentioned the highest. If the user is not sufficient for the optimization of the database itself and the application optimization, then the DB2 will be prone to lock-wait phenomenon.
In addition, DB2 is not perfect in the provision of APIs and functions. But people in the Toronto lab are trying to perfect the module, and in DB2 9 we expect to see more powerful functions.
At the same time, high availability for all database systems is a more troubling issue. High availability in DB2 may be more complex for average users, but IBM has some classic technical documentation that can be used to refer to helping users set themselves up. For DBAs who manage multiple partitioning systems, it is important to coordinate the partitions and how to properly set the partitioning key.
As a technical support group, it is important to have a truly in-depth understanding of the communication and working mechanisms between the various partitions. I remember reading an article that said DB2 share nothing architecture is a deceptive thing, personally think because he is the DB2 theory of multiple partitions just do some textual understanding, but for the memory CPU, the allocation of disk resources is not understood.
Perhaps many people think that DB2 is a lot more complicated than other database systems, but the individual believes that it is due to their lack of understanding of the underlying architecture and models. The real understanding of DB2 will reveal that each of these modules has been carefully designed. If you can optimize the system in a complete way, many common problems with the database will not occur.