Configuring data’distribution: selection and transformation, creating intelligent agents
Configuring data’distribution is quick, easy and scalable. You can thus:
Configuring data’distribution and creating intelligent agents
The data in a company must have the capacity to evolve. This means that it must be possible to trigger an action as soon as any data appears, disappears, changes value or takes a critical value.
For example, as soon as a new purchase order is entered in the system, an e-mail acknowledgement must be issued. Or as soon as a purchase order exceeding a given amount is entered, a warning message must be issued immediately to the sales director.
The creation of these intelligent agents is integrated in the configuration of data'distribution. Since data'distribution can call any existing program, you can define the required level of responsiveness without having to intervene on the installed applications.
Scalable configuration, simplified maintenance
You can configure data'distribution from any Windows machine with a LAN or remote connection (and even via the Internet). There are no limits on the number of machines that can be configured.
The parameters are configured and stored on the source machine. The target machine automatically receives the source machine's parameters. Parameters can be modified while replications are active; these modifications will be applied when you restart the machine.
The test and production environment concepts are also integrated; automatic processes can be triggered when changing over from the test environment to the production environment. data'distribution defines replication models so that a single configuration can be used and distributed to several targets. With overrides (user-defined default values), you can define a scalable, easy-to-maintain configuration. The configuration procedure has been specially designed to be extremely fast and scalable.
Configuration structured by environment
The configuration is structured by environment, which means you can separate the different applications covered by data'distribution.
For example, one environment for the backup is configured and monitored by the operation team, one environment for loading the BI system is configured and monitored by the BI team, one environment used for synchronizing an e-business is configured and monitored by another team, one environment for inter-application synchronization, and so on.
Selection and transformation configuration
Context-sensitive help for writing and configuring the selection/transformation script makes this phase fast and straightforward. Simply specify the structure of the required target and the procedure to follow for assigning or calculating the target data based on the source data.
Selections can refer to all or part of the source record, its previous values and the context, i.e. the operation type (addition, modification, deletion), the operator (user, program, job) and the time factor (date, time to the nearest millisecond). This means, for example, that you can take account of deletions - except those performed by the year-end purge program - to maintain a more comprehensive history in the data warehouse than that on the production machine.
The context can be modified to transform a physical deletion performed on the production machine into a logical deletion (modification) in the data warehouse, or to obtain a table that traces every operation performed on a given source table.
Calling user programs
Being able to call user programs means you can call your programs to reuse existing routines, perform selections based on complex rules, or recover information in a different format.
Since data'distribution can also be used to trigger transformation scripts and run programs on the target machine, you can decide where you want the processing to take place.
For example: data’distribution can be configured to trigger the recalculation of a multi-dimensional cube right after the data updated in a batch on the source machine has been replicated on the target machine.
data’distribution integrates a replication function in non-connected mode (broadcast) to load data in deferred mode on non-connected target machines such as those used by shops, a sales force or partners.
The data’distribution repository is open and documented, and can be accessed using third-party tools to perform consolidations, checks, an OEM integration or impact analyses.