on 09-21-2011 8:30 PM
Hi,
whats the ultimate fastests solution:
Reporting from DSO
or
Reporting from InfoCube
???
...same Data ?
Best
Martin
Hi
sorry,
-much theory ( it should,...)
-no facts.
Of course if i put data aggregated from a dso to a cube than its faster, less datasets more speed.
...but cubes pull much more data stored in DIM tables ...
Thank You
Martin
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Martin,
You're getting a lot of theory and no facts because there is no "ultimate fastest solution" for reporting. There is no right answer to the question. It depends on the type of data and on the type of query.
The rule of thumb is that for queries that return a small number of lines (definitely 1 lines, maybe up to 10s or 100s) and are restricted to a specific set up keys, a DSO with an index on the query key will give better performance. If you are doing queries with unpredictable filters and drill-downs, or queries that aggregate a large number of records, then a cube will usually give better performance.
Of course, we could come up with specific cases that would break these rules-of-thumb. So if you want a simple answer you'll need to explain what the dataset and query pattern looks like
Cheers,
Ethan
Hi,
The main difference in Reporting from Info cube and ODS is that InfoCube is meant for Multi-Dimensional reporting while ODS reporting is like 2 dimensional reporting
Refer the below link for info
Thanks
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Martin,
For the same volume and same kind of data, reporting is much faster and also most recommended to be done from cube.
Regards,
RahulM
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Martin,
Its based up on many criteria, Like volume of data .... what kind of report to be evaluated... etc..
As you aware of the designing technique of data base is Multi Dimension Modelling for data warehousing. And Purpose of cube to have analytical report not operational or details level of report.
Cube is having multiple dimension and you can do your analysis maximum upto the interaction of 3224 Masterdata.
In simple, when you want to go for detail level reporting or operational report you can do it in DSO level. It may be faster than cube based up to the volume of the records.
An analytical report or historical report would be very much faster in cube compare to DSO .
-
Thanks
BVR
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi,
If you look initial performance of report performance on DSO it is very fast compared to cube. But as data increases it will be slower. It again depends on data volume that DSO or IC contains.
If you are creating DSO report see SIDs are generated for better performance.
IC has many advantages like like compression,aggregates,BIA index etc and so on.With IC on top of DSO you are increasing volume or DB growth but report performance will be much better.
Thanks and regards
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
it will depend on the data volume.
Huge data volume - Cube will the obvious choice.
Less data volueme - DSO an be used as well.
with Cube, you can use BWA as well.
Regards,
Gaurav
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Of course it is reporting based on cubes. Imagine within DSO you have just flat structure=pure database table. Within cube you have data modeled in star schema=fact table surrounded by its dimensions.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
88 | |
10 | |
10 | |
9 | |
7 | |
7 | |
6 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.