Previous feature | All features | Next feature

SUBSYSTEM TRACING CHANGES IN DATABASES

Tracing updates in information systems may be problematic especially in geographically distributed networks. As an example we bring a three level structure, though the number of levels can be bigger. We nominally name the levels as ‘center’, ‘regional level’ and ‘district level’.

Additional difficulty that may arise can be the following. Different providers may have developed the bases containing data at different time. In order to implement the mechanism for tracing updates we would have to find all the numerous developers of the numerous bases and coordinate the actions with them and in the end bring in the necessary changes into the program. The updates also have to be carried out automatically.

After updating only the updated information should be sent. Sending all the base would take too much time.

UPDATING

Updating data is one of the most important aspects of the system operation. The process of receiving the changed data can be split into several stages:

• Receiving a current scan of data from the databases server
• Picking out the modified data
• Update package formation

The data collecting district server initiates the process of updating on the level of a district. The following are the structural parts of the data collecting district server:

• Server Index
• Server Archives
• Archives of differences

Server index contains information on data stored in the databases. Using server index we can say for sure if there were changes introduced to certain fields in the database.

Server archives is used for storing the current scan for data received from databases. The archive always has unique conformity to the server index and represents the scan for latest update.

The archives of differences represents the structure with information on the changes introduced from the last update process.

After finalizing the new data receiving process all the data on the new, updated and deleted notes.

In order to receive server updates it creates a new server archives. Further on in the process of receiving new updated data the server goes through the data in the databases for receiving data and checks for availability of the data in the server index (algorithms of comparison with data caching are used for comparing). When going through the data in databases the data is put into the new server archives and marking is carried out in the old server archives. In case of finding a new or changed note a message on that is put in the archives of changes. After finishing the process of checking the server goes through the old archives in order to find detached notes and adds data on the found detached notes into the archives of differences.

After finishing the process of updating the server has a certain number of archives of differences on the number of archives on the differences in the number of databases, which were set for indexing. The received archives of differences are sent to the regional center.