Regarding deleting index

Hi Friends,
If I delete the whole content of an Infocube and then do the reload,is it require to delete the index seperately.
If I don't delete the index how is it going to impact the load?
Regards,
Debjani...

It doesn't mean that Indexes will delete permanently .. Structure will remain same!
When you delete data from cube it also deletes the generated indexes which are existed .. there is no point of having indexes on deleted records .. it's logical
even it doesn't hold any data it allows you to create index and generate stats(ex. create a dummy cube and try to generate)
For the moment I don't have any document for it.. I will let know
For your 2nd ques.
Only for huge loads droping index will be usefull which will improve the loading performance!
For daily delta load with less volume of data you need not to drop Indexes .. if you do so it will take lot of time to rebuild the indexes(on whole data)

Similar Messages

Regarding "Delete Index" process in the process chain.

Hi Gurus
In the process chain, I have Delete Index -> Load Info pacakge -> Create Index process in the above order.
I am loading few records, so no need to delete indexes everytime. Can i remove the "Delete Index" process from the Chain without deleting the Create Index process. Or do I have to delete both.
Could you please clarify my doubt if possible in detail.
Thanks,
Regards,
aarthi

With the numbers you provided, you probably don't need to have the steps. With Oracle (not sure about all the other DB flavors) there are some other considerations:
Not having the indexes when loading the data can improve the load time since the bitmap indexes are not very efficient with respect to inserts/updates. The more dimensions you have, the more indexes there will be and the more noticable the impact of having the indexes present when loading.
The drop index process only drops the F fact table indexes. If you compress your cubes regularly so that you don't have many uncompressed requests, the index rebuild time will remain small, but if you have many uncompressed requests in your cube, the index rebuild time can begin to exceed whatever time you might save on the load with indexes deleted.
With bitmap indexes present, you can also occasionally receive a ORA-0600 deadlock error during the load process, and that can be prevented by dropping the indexes before loading or chose the load option packet by packet so that two packets are not trying to update the same index block at the same time.
Another concern in shops where reporting on the cube might occur during the load process - if you drop indexes prior to a load, any one trying to run a query on teh cube could have poor query performance since all the indexes will be missing onthe F fact table, this agina becomes more apparent the more data you have in the uncompressed fact table.
So it really comes down to your environment, but certainly drop the indexes any time you have large loads.

Delete Index in Process Chain Takes long time after SAP BI 7.0 SP 27

After upgrading to SAP BI 7.0 SP 27 Delete index Process & Create index process in Process chain takes long time.
For example : Delete index for 0SD_C03 takes around 55 minutes.
Before SP upgrade it takes around 2 minutes to delete index from 0SD_C03.
Regards
Madhu P Menon

Hi,
Normally index creation or deletion can take long time in case your database statistics are not updated properly, so can check stat after your data loading is completed and index generation is done, Do creation of database statistics.
Then try to recheck ...
Regards,
Satya

Error in Process chain while deleting index and create index

While executing the process chain for delete index the process chain failed with the below error message as "The system has not finished loading request 445315 into data target ZECC_C01". While creating index also it failed with error message as "Lock NOT set for: Generate index, RSSM_PROCESS_INDEX terminated, because InfoCube/DSO object ZECC_C01 could not be locked".
Please help me in solving this issue.
Thanks & Regards,
Bharathi

Hi,
Manage the data target - ZECC_C01 and check the request number 445315. Also check the subsequent data targets if further loading was erroneous. Check whether the request is green / Properly loaded or not.
You can also use RSBKREQUEST table to check the request and its corresponding data target.
This request is either red or yellow caused this failure.
Hope this helps.
Regards,

Delete Indexes Variant ended with errors in Process chain

Hi experts,
I have a process chain which executes dialy to load data Info Cube we create the index and finally we delete the index after that we have few variants..
Remaining all variants are executed with out errors except delete index variant I checked the error message it shows that <b>" The system has not finished loading request APO_R46PWZHRIA7JS9NOM8X879QEZJ into data target ZUP_CARF1"</b>
It is so happened weekly twice or trice not every day. one day it was running with out any errors and one day this process chan was ended with errors.
I need to know the reason why this variant was executed with errors.?
I need your valule sugestions regarding this issue.
Regards
sailekha

hi sailekha,
therevare some functional constraint of processes.
while deleting indexes
it is not possible to:
                              loading
                              indexes are being built
                              stastics are being built
                              during rollup
                              while compressing
                              while updating.
my be u can check which is the other process chain that does all above oprations in concerned infoprovider.
bye

Taking more than 30 min for Delete Index

Hi Experts,
I am trying to delete Index for one of my cube thorugh Process Chain, but every time it is taking morethan 30 mins, even we dont have much data in the cube.
Job Details:
Job started
Step 001 started (program RSPROCESS, variant &0000000003309, user ID ALEREMOTE)
Start process DROPINDEX ZCSDC303_DEL_IDX in run DBWIM5VFE7YNJOOGV1YK6KTCQ of chain ZCSD_PWR_CONS_CON
please help me out on this
Thanks in advance
David

Hello David
As you said you have very less data in that cube that means you are not loading too much data on that cube on regular basis.
Deletion and recreation is not absolute necessity. You just need to check if loading without deleting index is not very slow.
Then just drop this steps of deleting index..you can save your 30mins of deletion and recreation.
by the way...what is your database? In MS SQL for some cubes we do not delete index before loading data, which saves lot of time.
You can also check database statistics...and please do a RSRV check for this cube to check if there is any inconsistency,
Regards
Anindya

How to create process variant for delete index in a process chain

Respected all
I am creating a process chain, but unable to create process variant for the delete index. kindly tell me the
step by step proceure for creating the variant for delete index. also if we use variant which is previously prepared then will it be safe? will it affect the other running process chian.
pls reply
thanks
abhay

Hi,
Please do not use an already created variant for index deletion. It might delete indices of some other cube.
You can follow the following steps -
1. Open the chain in Edit mode and click on "Process Types" button
2. On the Left hand side you will have various process types. Choose "Data Target Administration" and expand it.
3. Choose the first process "Delete Index" (it will be marked with a trash bin sign) and drag it to your chain
4. It will ask you create a variant. Press the create Button
5. In new window enter the Process Variant technical name and description
6. Then in new window choose Object Type = Cube through dropdown and Object name via browsing. After choosing the cube name click on "Transfer Selections"
7. Save and return to your chain
8. It will automatically generate a create index step also after delete index.
9. You need to break the link between create and delete step and insert infopackage in between to get the following steps - delete index --> load cube --> create index.
Please let me know if this is helpful..
Regards
nishant

Issue with Delete Index in process chain

Hi All,
I have a situation with Delete Index in my process chain.
Delete Index step, before loading data to cube, is taking around 2Hr to complete but some times it completes within 5 mins.
Anyone faces such issue? If yes, how can I fix it?
Thanks and Regards,
Alok

Hi,
If Statistics are not maintained for the cube then also it might casue the high run time in index deletion.
if its urgent then, can u try doing it manually form performance tab....
or once process chain step is complete, create statistics for the cube. Next run should take less time
Please try.
Regards,
Gaurav

Pc-delete index

Dear experts,
Is there an oss note regarding the significance of deleting indexes before loadind data to say a Data Target. Why is it so imperative.
Thanks
Niraj

Hi,
It is part of best practices to drop indexes of the cube before loading data. It will not throw error evenif you don't drop the indexes. But if you drop the indexes loading time will be reduced as it will not try to build indexes while loading. So best practice is drop index - load data - create index.
Regards
GIthen

HOW TO DELETE INDEXES IN SE37

Hi all
This is Sudhakar
can any one help me out How to delete the indexes in SE37
Please tell me the steps
i will give the points also It is very Urgent
Thanks and Regards
Sudhakar
[email protected]

Hi Sudhakar,
If you want to delete the indexes in SE37 it is a screen for Function Builder and you can drop indexes with the function module and compress rebuild indexes too.
if you are using any FM then fine can you tell what is that FM you are using.
you can do it by going to RSA1 right click on cube and Manage --> Performance->
check indexes>delete indexes>repair indexes
and to compress goto Rollup -->Execute (check for the request id)
Thanks
Shankar

Problem with delete index of cube in process chain

HI ,i have a problem with my process chain,here on process name as delet index is ther which has been turned to red.when iam going inside of that in Display message iam getting the msg as"Indexes for InfoCube SMS_C06 have been deleted successfully".
since the process is in red next process is not been triggering,but i have to continue the process chain.
can anybody put some light into it.its very urgent........

Hi,
I ran into the same situation with one of my process chain that is part of a metachain. The process chain has the Delete Index that turned red. But the messages show that the indexes on all the cubes were deleted sucessfully.
As suggested, I did a repeat on the delete index variant, but no luck as it turned red again.
Anyone aware familiar with this issue?
Thanks,
Briston

SRMO - Cannot delete Index Category after System Copy

Hi,
I am trying to delete incorrect Index Categories using t-code SRMO after a system copy (restore db from backup) from our production system to our sandbox system.
When I select the Index Category and click on the "Delete Category" button I get this message:
Error during deletion of language 'EN'. Search
engine result '3147'
When I click on the Help button I get this:
Error during deletion of language 'EN'. Search engine result '3147'
Message no. SRET148
Diagnosis
When the index category is deleted, all languages, and consequently all physical indices on the search server (RFC server) are deleted.
If a physical index could not be deleted, the index category is not deleted or set to inactive either. This is to prevent the index from staying on the search server without R/3 inforamtion.
Procedure
Firstly clean up all languages / indexes for which this information appears.
Use the function to delete an index without R/3 information (if such an index exists) and / or the function to delete the language without an index from the R/3 System.
You can then call the function 'Delete index category' again.
Any ideas on what function they are talking about in the second to last paragraph: "function to delete an index without r/3 information"?
Thank you,
Neil

I have exactly the same problem, only that the index category still can't be deleted even after I deleted the languages first. I still get the error message '3147'. It says that there was an error deleting the languages. As procedure it is suggested (as before) to first delete the languages and then again the index category. But the languages are no longer displayed in the index category.
Does anybody know how I should proceed in order to delete the old invalid index and to create a new one for indexing with the new TREX?

Delete Request in Process Chain - still need to delete indexes?

I have a cube here that is a full load every day. In the process chain I delete the 1 request (because there's only one request in the cube at any time) and then I reload using a full load.
My question is that with a normal load into a cube I delete indexes and then load and then recreate indexes. Now with my cube since I am deleting the contents of the cube do I need to add the delete indexes process in the chain?
Do the indexes get deleted when the data is deleted?

Hello,
Yes, when you delete the content of the InfoCube the Index associated with also gets deleted. So in the process chain Delete Content -> Load Data -> Create Index will be enough.
Thanks
Chandran

Process chain delete index from TABLE

is it possible to put a step in a process chain (/rspc) to delete and rebuild indexes from a normal table?
I can only find a way to do this for cubes... ?
thanks!

Hi,
You have to export the table name and index name using the following str to FM :RSDU_DROP_INDEX
begin of rsdu_s_index,
    tabname              like ddstorage-tabname,
    indexname            type rsdu_indexname,
    dbidxname            type rsdu_index,
    fieldname            type rsd_fieldnm,    "like dd17v-fieldname,
    type_expected        type rsdu_indextype,
    type                 type rsdu_indextype,
    type_check           type rsdu_check,
    unique_expected      type rsdu_indexunique,
    unique               type rsdu_indexunique,
    unique_check         type rsdu_check,
    partitioned_expected type rs_bool,
    partitioned          type rs_bool,
    partitioned_check    type rsdu_check,
    status               type rsdu_indexstatus,
    status_check         type rsdu_check,
end of rsdu_s_index,
so, you need to pass table name and index name through this structure.
hope it help....
Regards,
Raju

Deleting Indexes Job failing with ORA-01418

Hi,
While deleting the indexes to the Infocube the job is failing with an Error: ORA-01418. In SM37 it is throwing the message SQL-Error ORA-01418 specfied Index does not exist.
Checked in RSRV for the InfoCube it is showing red while we execute for the repair.
Please advice... right replies will be appreciated.
Regards,
DMK

For BW from 7.0 you can take a look to note 1003360.
Have you tried the autorepair from rsrv?

Regarding deleting index

Similar Messages

Maybe you are looking for