Understanding Timestamp Propagation
When calculating the timestamp for a calculation, the calculation will propagate from the dependency to the dependent. For example, if calculation X is dependent on Y and Y is updated at time t1, then X's timestamp will also be t1.
Where a calculation has multiple dependencies, the calculations timestamp will always be the timestamp of the most recently updated dependency.
Understanding Missing Dependencies
A missing dependency can occur under three conditions:
• When a calculation references another calculation that does not exist or has been deleted.
• When a calculation references a SCADA measurement that does not exist or has been deleted.
• When a calculation references another calculation with a parse error.
If a calculation has a missing dependency, then it will be assigned the Missing Dependency quality, and its Error will indicate which reference is the cause of the missing dependency. For example, if calculation Y refers to a non-existent calculation Z, then Y's error is set to Missing Dependencies [Z] indicating that the cause of the missing dependency is Z in Y's calculation.
If a calculation has multiple missing dependencies then each one is displayed in the Error. For example, if calculation X references non-existent calculations Y and Z, then X's error is set to Missing Dependencies [Y],[Z].
If a calculation references another calculation that in turn has a missing dependency, then all calculations in the chain are deemed to have a missing dependency. The Error of each calculation will identify the complete chain. For example, for calculations A, B, and C:
Calculation
Definition
Quality
Error
C
[D]
Missing Dependency
Missing dependencies [D]
B
[C]
Missing Dependency
Missing dependencies [C][D]
A
[B]
Missing Dependency
Missing dependencies [B][C][D]