Partitioned a cube

Hi
I have partitioned a cube on 0CALMONTH, now i want to see the partitioned data in cube, how the data is seen in the cube, can we see physically the data as it was partitioned by 0CALMONTH or partition is only logical (not physically seen)? is there any separate table to see the partitioned cube data?
why do we do partitionng....

Hi Ramesh,
Once the F table is compressed ,data goes from F table to E table then u can no longer see data in F table.
You can control whether all the Requests in the F fact table get compressed or limit it up to a cetain specific request, a certain number of requests, or request over xx days old. Generally, you never want to compress all Requests, since you might need to back out a Reqeust, and once you have compressed a Request, you can't bak it out.
Good practice would be to verify the correctness of the data in a Request before you compress it.
When u partition ,u do it for the infocube based on two characteristics OCALMONTH or OFISCPER.The fact table is created on the database with one of the number of partitions corresponding to the value range. You can set the value range yourself.
keep in mind when you partition in RSA1, that you are configuring the partitioning for the E fact table (F fact table is already automatically partitioned by BW).
but key pt is that no data gets into this E fact table you configured for partitioning UNLESS you run the cube compression/collapse!
see the below link for more on partitioning.
http://help.sap.com/saphelpsem40bw/helpdata/en/e3/e60138fede083de10000009b38f8cf/frameset.htm_
If you need any other information, let me know.
Thanks and Regards,
Pavan Kumar Gali.

Similar Messages

Where are partitioning a cube based on calmonth & fiscper

Q1.where are partitioning a cube based on calmonth & fiscper ok In partitioning cube there are two types 1.physical partition & 2.logical partition so i want to know the both things and diferrences.

Hi Kishore,
Logical Partitioning means dividing the data into two or more cubes. The cubes will have the same structure, but will store mutually exclusive data sets. For eg. storing data from Calendar year 2000 to 2005 in one cube, 2006 to 2008 in another cube and 2009 to 2011 in another cube. As these three cubes will have the same structure, we can create a multiprovider to combine these cubes for comprehensive reporting. This can be done any characteristic....like company code, controlling area, sales organization, distribution channels or any time characteristics. This is more of data model designing option.
Physical Partitioning is done on one cube. This is based on the Calender Month or the Fiscal period. This physically partitions the Fact table of the cube. For eg. you want to partition a cube that stores 2 years of data. The maximum no. of partitions allowed here are 26... (24 for each of the 24 months and two for data not belonging to these 24 months - these two partitions are system defaults and we have not control over). However, we can define any number less that 26 and the system calculates the maximum number to be utilized. Supposing we specify, a maximum of 18 partitions to be used, the system will use a 14 partitions only - 12 partitions for the 24 months (each 2 months) and 2 on either side of the time scale.
rgds
Naga

Logical Partition of Cube

Hi BW experts,
Can anyone explain the steps to create logical partitioning of Cube data?
I want to do it per fiscal year. If we create 05 cubes(Same structure) each for say fiscal year 04' 05'...08'
now how will the data flow into respective cubes.
Kindly explain the steps.
thanks,
Nipun Sharma
P.S. points will be rewarded.

Hi Nipun,
1.Partitioning the cube in both physical and logical way helps you to increase its performance on Queriying.
Partitioning InfoCubes Using the Characteristic 0FISCPER (At Infocube maintainance)
Prerequisites
When partitioning using 0FISCPER values, values are calculated within the partitioning interval that you specified in the InfoCube maintenance. To do this, the value for 0FISCVARNT must be known at the time of partitioning; it must be set to constant.
Procedure
1. The InfoCube maintenance is displayed. Set the value for the 0FISCVARNT characteristic to constant. Carry out the following steps:
a. Choose the Time Characteristics tab page.
b. In the context menu of the dimension folder, choose Object specific InfoObject properties.
c. Specify a constant for the characteristic 0FISCVARNT. Choose Continue.
2. Choose Extras -->DB Performance --> Partitioning. The Determine Partitioning Conditions dialog box appears. You can now select the 0FISCPER characteristic under Slctn. Choose Continue.
3. The Value Range (Partitioning Condition) dialog box appears. Enter the required data.
Pls chk the below link for partioning..
http://help.sap.com/saphelp_bw33/helpdata/en/33/dc2038aa3bcd23e10000009b38f8cf/content.htm
http://help.sap.com/saphelp_nw2004s/helpdata/en/0a/cd6e3a30aac013e10000000a114084/frameset.htm
Re: logical Vs physical partitions ?
Regarding Partitioning
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/ce7fb368-0601-0010-64ba-fadc985a1f94
/message/4204952#4204952 [original link is broken]
How i can partioning the BW?
Partioning and ETable
What is the use of cube partition?
*pls assign points,if info is useful*
Regards
CSM reddy

Partitioning of cube

Gurus,
I am extracting AR data using standard datasources in BW3.5 and would like to partition the data by creating year specific cubes. I have to store data for around 15 years so I was thinking of having cubes with data for 3 years each and report using a multiprovider. But wanted to clarify a few things -
The data that will be coming for deltas, should that be going to the current cube or should my update rules be connected to all the cubes? What happends in case an older invoice is cleared, say from 1998 is cleared should the data be updated in 1998 cube as well as current or only in the current cube?
The company has 16 posting periods. If I am doing logical partioning based on 0FISCPER so should that partioning be from say 001/1998-016/2001 or will it be 001/1998-012/2001.
The company is not having fiscal variant based on calendar year, but is from April to March. Could this be a problem for me to partition the cube.
Thanks for your help.
Regards,
Venkat

Hi,
The data that will be coming for deltas, should that be going to the current cube or should my update rules be connected to all the cubes? What happends in case an older invoice is cleared, say from 1998 is cleared should the data be updated in 1998 cube as well as current or only in the current cube?
--> Chances of getting such old records are rare or almost ZERO.
--> Those records supposed to load into that particular cube only. Not to current cube(2009).
The company has 16 posting periods. If I am doing logical partioning based on 0FISCPER so should that partioning be from say 001/1998-016/2001 or will it be 001/1998-012/2001.
The company is not having fiscal variant based on calendar year, but is from April to March. Could this be a problem for me to partition the cube.
--> Fiscal variant required to report against fiscal poriod. To restrict at cube level.
Loading into a multiple cubes only useful, if query contains selection and filter restriction on infocubes. Or again accessing all cubes through multiprovider(multiple joins) creates Negative performance than a single cube.
Srini

Partition a cube, Jump queries.

Hi all,
1. When do we partition a cube?
2. Why should we link reports in jump queries? Where do we use them?
Thanks in advance,
Sameer.

Hi,
For Partition of cube
Just right click on Infocube -> Change or doble click on the infocube to go to infocube maintenace. From the menu choose Extras ->Partioning.
Now ,if say, your requirement is to extract data as it is from different tables, you can go with view. But if there is some complex business logic to be included, you can do that in a FM.
Check the folowing links for jumping.
http://help.sap.com/saphelp_erp2005/helpdata/en/81/941139a772b30de10000000a11405a/frameset.htm
http://help.sap.com/saphelp_erp2005/helpdata/en/5a/ec1e4203fadb2ce10000000a1550b0/frameset.htm
Please reward.

Partitioning of cube using Fiscal period

Hai Gurus
Gud Morning
Can u tell me how to partition a cube using 0fiscper ie Fiscal Year/Period.
I am having one in this
How to give value for 0FISCVARNT.
I just need the steps to partition the cube
Please suggest me.
Awaiting all ur replies.
Thanks
mahi

Hi Mahi,
To partition the cube double click the cube this will take you to the definition of the cube. Click on the extras button at the top of the screen. you will get a dropdown, there you can see Partition option, click on that you will get another window there you all time chars select on which you want to partition and continue.
you will get another popup window there you have to give Range.
When you activate the cube fact tble will be created in DB with number of partitions corresponding to your Range.
While you do this there should be no data in the cube.
Regards
Ananth

Logical Data Partitioning Info cube

Hi,
We have a Legacy system data in BW which is about 25 million reports span across 6 years and by Fiscal year/period.
Currently data is held in 6 DSO's and its one time load to cube...data load won't happen on daily basis as legacy system is de commissioned.
How to Logical Data Partitioning Info cube
Thanks

Hi Choudary,
One way is as follows: Convert the update process (for 0VENDOR) to flexible update. For this you will need to make 0VENDOR an InfoProvier (RDS1 > Master data tab). Then you need to create update rules from the 0VENDOR_ATTR InfoSource to 0VENDOR InfoObject. Since you now have the 0VENDOR_ATTR InfoSource you can also create update rules from this InfoSource to your cube and load the data.
Hope this helps...

Creating Custom Partitions on Cube

Hi all
We are running BW7 (2004s) over Oracle.
We are contemplating creating our own partitions on the EFact table of a cube. The reason for this is that currently the partitioning of cubes is only possible using time characteristics (0FICPER and 0CALMONTH).
We have a need to partition by a characteristic that is an 8 digit number.
Hence, we are proposing to create our own range partitions on this Characteristic.
My questions are as follows:
1. If we do this, we would be creating them directly at the Oracle level. To do so, we would have to first create a copy of the Fact table definin then partitions. We would then drop the orginal fact table, and rename the copied table back to the original efact table name.
2. Are there any issues with us doing this? i.e. if for some reason the table is dropped in the future (for some unknown internal BW database process), would we lose the partitions?
3. We are keen to do this so that queries which are keyed by this 8 digit number are much more efficient. Also, when deleting data, we would delete by a partition (i.e. much faster to delete).
Has anyone else done this sort of thing? Any issues encountered? Any advice?
Thanks
Adam

I face the opposite problem, nobody wants to get rid of anything, so we jump thru hoops trying to figure out how to keep the data in BW and still maintain query performance. I'd love a few clients that want to get rid of stuff, perhaps we can trade a few folks for a couple of months.
You didn't say if this 8 digit char is a line item dimension, or if there are other chars in the dimension. That has query and deletion performance implications.
You can partition your ODSs on the 8 digit char.
On the cube side side, I have never explored it further, but wondered if I wasn't using 0FISCPER, (or it could be 0CALMONTH) could I populate it in an update rule with some date values that could serve as the proxy for the characteristic ( or ranges of values in your case ) I really wanted to partition on, e.g.
8 digit char
00000001 - 00999999 = 0FISCPER 1000/000
01000000 - 01999999 = 0FISCPER 1000/001
etc.
I think you can probably get the idea. I think this would support the goal of the faster Selective Deletion, but obviously doesn't help you on the query side - although you might be able to have your user exit translate the 8 digit char to the corresponding 0FISCPER proxy. Seems kind of goofy, but I think it could work, but ultimately might cause more confusion than it's worth.
Query performance would only benefit from partitioning, of course, if the queries would restrict based on the partitioning characteristic to make Oracle prune the unneeded partitions from the query.
<u>Selective Deletion</u>
As I mentioned in my first post, Selective Deletion does have some performance tweaks built into the process besides dropping the partition. If the deletion was going to delete more than 10% (the deletion process figures this out) of the rows (I don't remember if you can control that % by an RSADMIN parameter or not - although you could probably modify the program or create a Z version to use some lower %) it will create new E and F Fact tables, copying the data to the new tables, omitting the rows that meet your deletion criteria, and allows you to specify how many parallel processes to launch. Once it's done, it deletes the original Fact tables and renames the new ones. You do need to have enough disk space to hold the original cube and the new, smaller cube, since they both exist for a period of time. This is much faster than the heavy lifting of Delete query, and could be run off hours and wouldn't require any special DDL or customization of the fact tables, or prevent use of aggregates, etc.
Or you could go the other route of running a daily deletion process to keep the deletion shorter. If the number of rows to be deleted on any one day is small and there is an index on the 8 digit Char, it might be fast enough to suit your purposes.
I'll ponder on this a while and see if anything else comes to mind. Some time back, I thought I had read a post that mentioned a custom partioning that had been developed for a DB2 customer. I followed it up with a Custosmer Msg to SAP recently to see if there were any other partitioning options besides the two dates for Oracle. I got a "not at this time" answer. If you haven't contacted SAP about this issue already, perhaps if you post a msg, they'll get a better sense of the level of interest.
If you're attending Tech Ed (SDN Day?) in Vegas in Sept, we could explore this further.
Pizzaman
PS.
Put yourself on the SDN world map (http://sdn.idizaai.be/sdn_world/sdn_world.html) and earn 25 points.
Spread the wor(l)d!

How to change the Mesure Group ID, without losing the existing partitions in cube developement

Hello all,
I am creating my cube and i created measure groups and partitions for those Measure Groups. The problem is now i want to change the measure group id. but when i do that, i am losing my partitions and aggregations on those partitions. is there way to change
the measure group id with out effecting/losing the partitions? please let me know your thoughts.
Thanks in advance.
Prajwal
prajwal kumar potula

Though its not my question, but just curious, where these references might be? What are we looking for?
Are you talking of permissions in Roles etc or its somewhere additional we might have references?
Not it will not be in permissions or roles. Off the top of my head you would be looking for references to "Measure Group1" in child objects of the measure group, so partitions and aggregations and maybe calculated measures. They would most likely look something
like the following:
<MeasureGroupID>Measure Group1</MeasureGroupID>
http://darren.gosbell.com - please mark correct answers

How to see Partition in Cube

Dear Gurus:
Can u pls tell how to partition cube on 0CALMONTH /FISCPER
Also, <b>how can we see the partition info on an existing cube...</b>
As always, thanks a lot in advance.
Best Regards

Go to the infocube maitenance screnn Extras->partioning.
Partioning is done based on the calender month(0CALMONTH) or fiscalyear/period(0FISCPER). At least one of the infoobjects must be contained in the cube.
This function suports only by the ORACLE, INFORMIX
Have a look also at this:
finding the partitions in a cube !
Regards

Partition on cube

Hi All,
We have some issue on Query performance.I want to know that Cubes on which this Query run has some partitioned or not how can i know this ?
<removed by moderator>
Thanks,
Deepak
Edited by: Siegfried Szameitat on Jun 22, 2011 7:43 AM

First of let me tell why we go for partition ,Partitioning is used to optimise read performance of queries.
For ex:-If you have 10 years of data in the infocube,but
you user requirement is only they want to see the data from
2 year.
Case 1:-Without partitioning whenever we run the query it will search
the relevent data from all the 10 years data,so here the performance of the
query will decreased as serach operation will take more time.
Case 2:-With partition For example if we have created partition
for 2006-2008 not system automatically creates two partition
ie before and after P1(Before partion):-2000-2006 P2:-2006-2008
P3 (After Parttition):-2008-2010.Now wheneever we run the query
it will directly go to the relevent partition and serach for the relevnet reord
here the query performance will be high as the serach operation will take
less time.
Physical partition: It is done when you want to partition the data in your cube in the database using 0calmonth or 0fiscyear.
Logical Partition: This method is used if your data in your cube become huge. So you create multiple cubes say for 1cube for 2001, another cube for 2002, etc... You can then use a multiprovider - which consists of all these cubes and then report on all the data.
You have to choose either 0FISCPER or 0CALMONTH wich is defined by the usage of your most-used big queries (i.e. weekly , monthly, quarterly) and calculate the number of necessary partitions by giving a time period from / to.
Make sure you get a good rows per partition ratio by keeping record numbers big in each partition and partition numbers small by giving a maximal partition number
i.e.
CALMONTH:
max. partition number = 96 (add 1 for MAXVALUE partition)
from 01.2000 to 12.2007 = 8 * 12 + 1 = 97 partitions (for each month 1 partition)
max. partition number = 24 (add 1 for MAXVALUE partition)
from 01.2000 to 12.2007 = 8 * 3 + 1 = 25 partitions (for each quarter 1 partition)
Thanks,
Saveen Kumar

How to Partition a Cube

Hi,
We are facing some performace issue with a cube. Can you please advise how we can analyze it and do partitioning on that. Please help for OLAP and cubes performance issues. The database is 11g on RedHat 4 x64.
Thanks
Naveed

Hi, Please give us some more detail on your model and data.
Dimensions, how many members, data volumes, update frequency, how many measures, any calculations?
Also details on the server spec.
11g - do you have the OLAP patch(es) installed?
What query tool will you use?
Thanks.

Re-Partitioning a cube

I have a cube that has 0CALMONTH data as follows, but no partitioning identified.
SQL
my_date               # Fact Records
200705                3
200712                3
200805                1
200808                2
200810                4
200811                28
200812                8
200901                20
200902                18
200904                16
200905                22
200906                29
200907                58
200908                67
200909                99
200910                57
200911                47
200912                63
201001                82
201002                193
201003                186
201004                237
201005                125
201006                392
201007                606
201008                1218
201009                1552
201010                2175
201011                3573
201012                2649
201101                5538
201102                12616
201103                34800
201104                91420
201105                1678650
201106                3118399
201107                1190442
201108                406521
201109                81457
201110                31661
201111                9305
201112                3561
201201                3933
201202                2439
201203                2694
201204                2098
201205                1980
201206                1747
<snip>
In fact this actually goes to year 2025.
My question then is do I partition by 0calmonth and create that many partitions?
I am partitioning using RSA1 -> Add'l Functions -> Repartitioning
Mike

Micheal,
You could partition the same but I think I vaguely remember that SAP BW restricts the number of partitions that you can create.
If BW does not allow you to create that many partitions - you could create partitions for the most used data and have the rest going into the max and min partitions - but then if you are extracting data from this cube - this approach would not work because if the max and min parttions increase in size considerably - then even if you specify a month which needs to be pulled - the exttract from the cube will do a full scan on the fact table.

Partitioning a cube in a Process chain

Hi,
How can I launch the partitioning of my cube in a process chain ?
Is there an ABAP program i can use with a variante ?
I have 3 cubes (one per Year : Y, Y-1, Y-2).
I would like to create a process chain for Shifting my data :
Delete data of cube Y-2
Load data Y-1 in cube Y-2
Delete data of cube Y-1
Load data in cube Y-1
Selective deletion in cube Y
When my cubes are empty I would like to make a new partitionning with the next year.
this | Next |
year    | year   |        old part.         |         new partioning
cube Y_           2007   | 2008      | 001.2007->012.2007 |      001.2008->012.2008
cube Y-1        2006   | 2007      | 001.2006->012.2006 |      001.2007->012.2007
cube Y-2        2005   | 2006      | 001.2005->012.2005 |      001.2006->012.2006
Thanks for any help.

Hi,
run transaction delete_facts and generate a deletion program for your cubes with a select-option for the year. Create variants for that programs to delete the relevant years.
Create a chain with the following steps:
1st. abap to delete year - 3 from last cube
2nd. load year - 2 from previous cube to deleted cube
3rd. abap to delete year - 2 from previous cube
4th. load year - 1 from actual cube.
5th. delete year - 1 from actual cube.
Hope this helps!
kind regards
Siggi

Is there a way of partitioning the data in the cubes

Hello BPC Experts,
we are currently running an Appset with 4 Applciations. Anyway two of these are getting really big.
In BPC for MS there is a way to partitioning the data as I saw in the How tos.
In NW Versions the BPC queries the Multiprovider. Is there a way to split the underlying Basis Cube to several (split by time or Legal Entity).
I think this would help to increase the speed a lot as data could be read in parallel.
Help is very much appreciated.
Daniel
Edited by: Daniel Schäfer on Feb 12, 2010 2:16 PM

Hi Daniel,
The short answer to your question is that, no, there is not a way to manually partition the infocubes at the BW level. The longer answer comes in several parts:
1. BW automatically partitions the underlying database tables for BPC cubes based on request ID, depending on the BW setting for the cube and the underlying database.
2. BW InfoCubes are very different from MS SQL server cubes (ROLAP approach in BW vs. MOLAP approach usually used in Analysis Services cubes). This results in BW cubes being a lot smaller, reads and writes being highly parallel, and no need for a large rollup operation if the underlying data changes. In other words, you probably wouldn't gain much from semantic partitioning of the BW cubes underlying BPC, except possibly in query performance, and only then if you have very high data volumes (>100 million records).
3. BWA is an option for very large cubes. It is expensive, but if you are talking 100s of millions of records you should probably consider it. It uses a completely different data model than ROLAP or MOLAP and it is highly partition-able, though this is transparent to the BW administrator.
4. In some circumstances it is useful to partition BW cubes. In the BW world, this is usually called "semantic partitioning". For example, you might want to partition cubes by company, time, or category. In BW this is currently supported through manually creating several basic cubes under a multiprovider. In BPC, this approach is not supported. It is highly recommended to not change the BPC-generated Infocubes or Queries in any way.
5. If you have determined that you really need to semantically partition to manage data volumes in BPC, the current best way is probably to have multiple BPC applications with identical dimensions. In other words, partition in the application layer instead of in the data layer.
Hopefully that's helpful to you.
Ethan

Partitioned a cube

Similar Messages

Maybe you are looking for