Data partitioning

Data partitioning enables you to control which data a user can see. For example, you can configure your system so that customers (end-users) can see only requests that they have logged, or so that analysts can see only those requests that are assigned to their current support group.

You can specify which data is partitioned and which isn't: for example, you could set up partitioning so that all incidents, problems, and changes are partitioned, but categories and CIs are not.

You can also specify for which customers or support groups data partitioning is enabled.

All types of user (Analyst, End User, Account Manager, and Contact) can be members of all types of group (Support Group, Company, Customer, and Supplier), and can switch their current group between these different types of group. This enables you to configure systems, for example, where an analyst is associated with a particular customer.

Because a user can be in different types of group, both types of data partitioning can potentially be seen by all types of user. The type of partitioning observed depends on the type of group a user is currently in. If an analyst switches to a customer group, they see customer partitioning; if they switch back to a support group, they see analyst partitioning.

By default, customers can see only data that is associated with themselves, and analysts can see only data that has a last assigned group that matches their current group.

However, you can define the related attribute that is used to define data partitioning for each User Type. For example, you could decide to allow analysts to see only data that is associated with their office, or allow end users to see only data that has a specific category.

An example of data partitioning

Consider the following set of requests:

Incident ID	Customer	Last Assigned Group	Office
1	Customer A	IT	UK
2	Customer B	IT	US
3	Customer C	Hardware	UK
4	Customer B	IT	US
5	Customer A	Hardware	UK

With no partitioning enabled, all analysts (and customers who are allowed to log in) can see all five requests.

With the default partitioning enabled, analysts whose current group is IT see only requests 1, 2, and 4. Customer B sees only requests 2 and 4.

However, if you use the Office attribute as the partitioning attribute, an analyst in the UK office sees only requests 1, 3, and 5 – irrespective of which support group they are in. Similarly analysts in the US office see only requests 2 and 4.

You do not have to enable partitioning for all groups – you can leave some groups using unpartitioned data. In this way you can design a system where some groups see all data, and some see only data that is directly relevant to them.