Logical data groups overview

When examining monitor element data, it is frequently useful to look at multiple related elements at the same time. Logical data groups are groupings of elements that complement each other. For example, the uow logical data group includes elements like appl_id (application ID) and appl_name (application name), which are two elements that you can easily imagine wanting to see together.

Logical data groups are used by snapshot monitoring interfaces, and in particular, by event monitors. There are over a thousand monitor elements in the DB2® product. When examining monitor elements, it would be very tiresome to be forced to always have to think about and be forced to specify specifically which elements you want to see. For example, if you are creating an event monitor, thinking about and specifying what elements to capture data for might be very tedious. Instead, the DB2 product associates a default set of logical data groups with each event monitor. This means that you do not need to specify anything in the CREATE EVENT MONITOR statement to capture a useful set of monitor elements; only monitor elements relevant to the events being captured are included. For event monitors that write to regular tables, you have the added flexibility of being able to specify the logical data groups for which you want monitor element data captured.

Event monitors that write to unformatted event (UE) tables capture also capture a default set of monitor elements; when you use the EVMON_FORMAT_UE_TO_TABLES procedure to generate relational tables, logical data groups are used to group related elements together in separate tables. For example, the lock logical data group contains the elements used in the LOCK_EVENT table; the participant logical data group contains the elements used in the LOCK_PARTICIPANT table.