238 IBM Cognos Dynamic Cubes
To generate some insight about which aggregate cubes are considered and why they are
selected, enable the logging event group specifically for the database aggregate
routing-related activity:
1. On the QueryService system, edit the DQM logging configuration, which is located relative
to the server installation directory:
configurationxqe.diagnosticlogging.xml
2. Edit the aggregate cache event group to log level of trace:
<eventGroup name=" ROLAPQuery.AggregateStrategy" level="trace"/>
3. Save the xqe.diagnosticlogging.xml file.
4. Restart the QueryService for the logging level changes to take effect.
5. Start the cube.
6. Run the problem query.
7. By default, the DQM server log is written to a file, which is located relative to the server
installation directory:
logsXQExqelog-<timestamp>.xml
8. When troubleshooting by using the DQM server log is complete, revert the changes to the
log level in the xqe.diagnosticlogging.xml file, and then restart the QueryService for the
changes to take effect.
As queries are processed in Cognos Dynamic Cubes, the analysis of which aggregate cubes
are considered, which are rejected, and which one is selected, gets written to the DQM server
log. A single user query may be decomposed into smaller queries so that some values that
can benefit from database aggregates are routed to an aggregate table and other values are
routed to the underlying warehouse tables.
The analyses for each of the smaller queries are found in log entries for the database
aggregate routing event group denoted by the <aggregateAnalysis> element. The aggregate
analysis element consists of the following sections:
? Original query: Lists the measures, levels, and each level’s dimension and hierarchy from
the input query for consideration.
? Aggregates considered: Lists all of the matching aggregates that are qualified for routing
the input query for consideration. There might be more than one matching aggregate cube
that is qualified and considered as a match.
? Aggregates selected: Describes the final aggregate that are selected from one or more
qualified aggregates for consideration. It also describes the reason for choosing this
aggregate.
? Aggregates not matching: All the aggregates that do not qualify for routing are listed here.
The reason for not qualifying is also described for each non-matching aggregate. Typically,
reasons are as follows:
– Measure mismatch: The input query measure aggregation is not listed as part of the
aggregate cube, and therefore cannot be used.
– Level mismatch: The levels at which the aggregate cube is defined do not match with
the level of aggregate of the input query.
– Aggregate slice mismatch: The slice coverage of the aggregate cube is not enough to
satisfy the input query even though the measure and levels are matching. For example,
if the aggregate cube is defined for slice of Time Year 2012, and the input query is