Query Logging Sample Output

The following seSample Query Log Outputgment shows an example of how log settings look in a log file. In the example, the log settings show that all members of Product are logged and that members of generation 2 of Market are logged by generation number. The log format is cluster and the log path is C:\QUERYLOG\.

  <?xml version="1.0" encoding="UTF-8" ?> 
- <root>
  - <session>
     <bootuptime>Wed Jul 23 15:27:26 2002</bootuptime> 
   - <logsettings>
    - <dimensions>
     - <logdim name="Product">
     - <logdim name="Market">
        <spec>GENERATION 2</spec> 
      </logdim>
     </dimensions>
  - <othersettings>
     <logformat>cluster</logformat> 
     <logpath>C:\QUERYLOG\</logpath> 
   </othersettings>
   </logsettings>

Description

A query is a unit of retrieval from the user perspective. The way a user may perceive a query is different than how the server analyzes and executes a query. Even if a user performs a single retrieval, in order for the server to efficiently execute the logical query, the server splits the query into a number of subqueries to execute. Therefore, a single retrieval from the user perspective may actually consist of several subqueries from the server perspective. These subqueries are reflected in the query log.

Sample Cluster Output

The following segment shows an example of how queries are logged in cluster format. The username is listed along with the query execution date and the start time of the query. Each cluster contains two dimension entries. The first cluster shows that members 100 and 200 of the Product dimension were queried. The second cluster shows that member 300 of Product and Generation 2 of Market were queried. The elapsed time to perform the query is also provided.

<query>
 <user>User1</user>
 <time>Tue Aug 13 12:29:49 2002</time>
 <subquery>
  <cluster size="2">
   <dim size="2">
    <member>100</member>
    <member>200</member>
   </dim>
   <dim size="1">
    <member>Market</member>
   </dim>
  </cluster>
 </subquery>
 <subquery>
 <cluster size="2">
  <dim size="1">
   <member>300</member>
  </dim>
  <dim size="2">
   <member>Market</member>
   <generation>2</generation>
  </dim>
 </cluster>
 </subquery>
 <elapsedtime>0.016 seconds</elapsedtime>
</query>

Sample Tuple Output

The following segment shows an example of how queries are logged in tuple format. The username is listed along with the query execution date and the start time of the query. Note that each member of Product is displayed with Market. Each possible member combination is displayed for a given query. The elapsed time to perform the query is also provided.

<query>
 <user>User1</user>
 <time>Tue Aug 13 12:28:14 2002</time>
 <subquery>
  <tuples>
   <tuple>
     <member>100</member>
     <member>Market</member>
   </tuple>
  </tuples>
 </subquery>
 <subquery>
  <tuples>
   <tuple>
    <member>200</member>
    <member>Market</member>
   </tuple>
  </tuples>
 </subquery>
 <elapsedtime>0.02 seconds</elapsedtime>
</query>