Thursday Mar 28, 2013

Fact Driven

The P6 Analytics product is driven by Facts. When you are looking at a subject area, take Primavera - Activity for example. You can add from the Project Dimension Project Id and Project Name. This will list out all projects in the w_project_d dimension table. In your Analysis if you add a Fact - such as actual cost - this will join the Project Dimension to the Activity Spread Fact table. When making this join a fact record must exist or you will get No result set found returned. You would have this situation if the project you are looking for either has no activities or does not have any spread records.

In P6 Analytics versions 1.2 and higher this could occur if the project has not been published in the Extended schema. If you know there are activities that should contain spread data the best place to begin evaluating would be the extended schema. In the P6 application go to the Projects section and add columns for enable publication (make sure it is checked) and last published (make sure this date is populated and recent). You can right click on the project in P6 and choose publish now. Check back in a few minutes and see if the last published date has been updated. More more information on the Extended schema check out the P6 Extended Schema white paper.

Friday Mar 22, 2013

P6 Reporting Database and additional Java options

When executing the ETL process for the P6 Reporting Database you will be running startetl.bat or .sh.  In this file there is a JAVA_INVOKER_PREFIX.
This will have pre-determined options that are needed by the ETL process.  If you need to include additional java options they can be added to this parameter.  One item that may need to be added is to ensure there is enough entropy on this server. If issues are encountered making a JDBC connection it could be related to the entropy on the machine. Validate that the server has enough entropy. If there are a low about about of entropy please contact system administrator for server. The ETL process needs to make both TNS and JDBC connections. 

Friday Mar 01, 2013

Changing Lines at the data date

Have you ever wanted to show actual values vs. remaining on the same line using different colors for each? This can be a very tricky exercise in OBI. First, you have to create 2 separate date sets, one set of dates that represent time before the data date for actual values, and another set of dates after the data date for whatever is remaining. After this, you want to make sure that you can clearly show the data date so it is known where the actual values end and where the remaining begin <see chart below>.  After you have created the time separation, you can begin to assemble your graph. In OBI, the order in which you place your lines matters for the simple fact of which bar will overlay the other. Since our actual bar will stop at the data date, we want this bar to "appear" on top of our remaining, so we will see actual values up to the data date, then remaining after the data date. The last part of this involves the display of the data date. I accomplished this by placing a large number in a field that I used for the data date. This value represents the height of the bar essentially, so when you try this in your environment, you may need to adjust the value according to your projects values.

If you want to see the analysis (OBI 11.1.1.6.7), download the files here


About

Provide new information on P6 Analytics and Reporting Database.

Search

Categories
Archives
« March 2013 »
SunMonTueWedThuFriSat
     
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
23
24
25
26
27
29
30
31
      
Today