So what should the outline order be for your BSO databases? Many of you Essbase experts have heard this before but just in case…
Outline ordering (and in general Essbase tuning) is not an exact science. But I recommend you start here and then tune / test …
- Largest Dense Dimensions
- Smallest Dense Dimensions
- Smallest Aggregating Sparse Dimensions
- Largest Aggregating Sparse Dimensions
- Non-aggregating Sparse Dimensions
Dense dimensions - define the data block and must reside at the top of the outline
Aggregating Sparse dimensions - dimensions that will be calculated to create new parent values
Non-Aggregating Sparse dimensions - dimensions that organizes the data into logical slices
- Placing these dimensions as the first Sparse dimensions positions them to be the first dimensions included in the calculator cache
- Data is often times more dispersed within the database
- Example - Scenario, Year or Version
Here’s an dimension ordering example based on member count:
Dimension | Type-Size |
Accounts | D – 94 |
Time Periods | D – 21 |
Metrics (Hrs, AHR, $) | D – 14 |
Scenarios | AS – 9 |
Job Code | AS – 1,524 |
Organization | AS – 2,304 |
Versions | NAS – 7 |
Years | NAS – 7 |
Here’s an example based on dimension density (Ordering the dense dimensions from most dense to least dense maximizes the clustering of thedata):
Dimension | Type-Size | Density After Calc | Density After Load | Data Points Created |
Time Periods | D – 21 | 85% | 85% | - |
Metrics (Hrs, AHR, $) | D – 14 | 22% | 22% | - |
Accounts | D – 94 | 3 % | 2% | - |
Scenarios | AS – 9 | 22% | 11% | 199 |
Job Code | AS – 1,524 | .56% | .23% | 853 |
Organization | AS – 2,304 | .34% | .09% | 783 |
Versions | NAS – 7 | 19% | 19% | - |
Years | NAS – 7 | 14% | 14% | - |
How do I find the density of each dimension?
- Make the dimension the lone Dense dimension
- Load and calculate just that dimension
- Check the block density value in Administration Services >Database<> Statistics
Another consideration for outline ordering is your compression type. One compression type RLE (Run Length Encoding) is a good compression type when your data has many zeros or often repeats (found in our budgeting applications. RLE will evaluate and use RLE, Bitmap or IVP for compression. So how does outline ordering impact compression? The first dense dimension determines your “columns” in PAG file. Compression will take place from left to right, top to bottom.
The standard method is to place Accounts / Measures first in every outline. If we were to do that, the .pag file would look like the following (simplified version):
BUDGET | Sales | COGS | Margin | Exp. | Profit |
January | 100 | 50 | 50 | 30 | 20 |
February | 100 | 50 | 50 | 30 | 20 |
March | 100 | 50 | 50 | 30 | 20 |
April | 120 | 50 | 70 | 30 | 40 |
May | 120 | 50 | 70 | 30 | 40 |
June | 120 | 50 | 70 | 30 | 40 |
But if we move Time to first dense dimension and we get the following (notice repeating values):
BUDGET | Jan | Feb | Mar | Apr | May | Jun |
Sales | 100 | 100 | 100 | 120 | 120 | 120 |
COGS | 50 | 50 | 50 | 50 | 50 | 50 |
Margin | 50 | 50 | 50 | 70 | 70 | 70 |
Exp. | 30 | 30 | 30 | 30 | 30 | 30 |
Profit | 20 | 20 | 20 | 40 | 40 | 40 |
So… Time should be dense and listed first in the outlinewhen using RLE compression.
2 comments:
Hello Tracy,
You mentioned after the compression topic that,
Accounts/Measures dimension should be first in the outline.
But, after the sample table, you wrote Time has to be the first dimension.
Did I interpret in incorrectly. Please explain..
Thank You
Its very useful to me. Wonderful blog.. Thanks for sharing informative Post.
Installment loans in alabama
Payday loans in alabama
Title loans in alabama
Cash Advances in alabama
Post a Comment