Name | Type | Multiplicity | Description |
---|---|---|---|
value |
property int |
[0,1] | Primary (absolute) value of that stat. |
domain |
property int |
[0,1] | Size of the domain. |
percentage |
property float |
[0,1] | Percentage value. |
timestamp |
property dateTime |
[0,1] | Sampling timestamp. |
Flags: RAM,runtime
Multiplicity: [0,1]
Primary (absolute) value of that stat. This is the value that we are interested in.
E.g. this is the number of objects in collection.
Even though this is a primary value, it is optional. There may be rare cases
when we do not have primary value, we only have percentage data.
Flags: RAM,runtime
Multiplicity: [0,1]
Size of the domain. This is the value of "all things" used to compute percentage.
E.g. this is the number of object in collection domain.
Flags: RAM,runtime
Multiplicity: [0,1]
Percentage value. This is usually not present, as it may be easily computed
from value and domain. But for rare cases there may be just the percentage and
the primary value may be missing.
Implementation note: Having percentage here also makes it easier for implementation.
The code can use the same data structure to store intermediate values, e.g.
stat with computed percentage. The collection-processing code can also fill
in the percentage value automatically, even if it is not stored in repository.
This is somehow similar to the way provisioning code fills in the attributes
into the shadow.
Flags: RAM,runtime
Multiplicity: [0,1]
Sampling timestamp. The timestamp of the moment when the value was
retrieved or computed.