|
Collaboration and Group Support
You will see the following criteria almost repeated in some other tool categories as these are of generic order. Any large enterprise with fast-track initiative will have multi-staff working on a task. Most of these collaboration and group support capabilities are like an add-on layers, which can sit on various tools (like a work-flow system), as long as you have integration hooks.
- Multi-user design environment supports collaboration on large, enterprise-wide projects: One should have a collaboration capability, with some inherent work-flow type environment, whereby the DQ configurations can be done by multiple users simultaneously. This will be supported by tracking of changes, resolution of conflicts and manual acceptance of changes.
- Ability to distribute data quality tasks: This is like work-distribution capability typically found in the work-flow based systems. One should be able to assign chunk of tasks to different people, manage resource loads and track the time taken to complete a task.
- Version management with concurrent developers and release managers: Version management with multiple users is a tricky task compared to a sequential check-out and check-in. However, from the user’s perspective it is not as there are many smart tools, which handle concurrent version change management.
- Central Library of data quality routines to enable the re-usability: A central library is like a metadata repository, which provides the features like search, grouping, tagging and detailed documentation. This will help the users to pick-up to look for and use the routines which suit them. This is a blessing for large companies.
Administration
These are again the standard features of a good enterprise level tool-set.
- A common console to track the data quality related activities: or link to console of other larger tool-set. You should have one administration control panel, which will allow you to manage all what you need to do for your DQ tool. This will actually not be a single screen (as all the DQ management related tasks cannot be shown in a single view). This panel will be an integrated display of all options and tasks, like wizards, scheduling management, DQ operations tracking, version management, user access management, metadata management and so on ...A DQ tool firstly should have it own central management facility. If possible, it might be also able to become part of the central management facility of a larger tool-set like a Data Integration Tool.
- Ability to define the job runs operations: This will be similar to any data centre operations of job flows
- Creating the entire operations flow for sequential and parallel activities
- Define the number of simultaneous threads
- Define the time caps for the batches to run
- Define the mandatory and optional batches
- Define on the activity logging linked to various batches...
- Automatic or manual management of the sequence of activities: You can set-up the entire operation in a auto-mode and to be stopped only if it hits certain conditions. Alternatively you can have it all manual or a combination of auto and manual.
- Manual or automatic validation checks, pre and post batches: This is the extension of the previous point. The system should be able to do checks on the auto as well as manual modes. It should be able to create alerts, messages and logs on the results of those checks.
- Work-flow based capability to do scheduling: The whole scheduling can be defined using a work-flow functionality where you can create components of schedule, and someone else can tie them all togather or approve the same.
|
