0HR_PY_1 Performance issues

[former_member182869]

Dear Experts,

Can any one Faced any issues while Extracting Data from 0HR_PY_1, we want to extract data from this extractor in production enviornment,

it will be huge data say' 6 to 7 years Data, want to analyis what kind of perfomance may going to come before moving to production system.

Cany any one have views on it.

Regards

Sree

[former_member182470 · Accepted solution]

This is a very generic question. Performance issues will arise for any thing in the system if you are not followed proper settings.

You have to load your data in chunks to finish all 7 years data. Increase the data transfer limits in SBIW in ECC system.

Also increase in your Infopackage)scheduler menu-->Default data transfer settings) and in RSCUSTV6 t-code.

Make sure you have proper RFC connection with all ports and services.

Please check below blog:-

[former_member182869 · Accepted solution]

Dear Suman,

Seems to be basic question in terms of LO Datasources, but in the case of Payroll datasources it will give a problem with data, we have used to more than 10 Datasources with same Function module for extracting huge data for one of the other datasource which is built on cluste tables.

if one have faced such problems in payroll, give their scenerios.

Regards

sree

[former_member182516 · Accepted solution]

Hari,

One issue strike my mind ... while payroll data  load is running in BW make sure no payroll postings are running in source. If payroll postings are in progress in ECC and your BW data load got triggered then the load will fail with lock issue .

error message:

Posting runs are locked. Restart the transfer later Errors in source system

This is one thing you can make a note of it... make sure you check with business what are the feasible times when you can run your BW data loads so that there won't be locking issues.

FYI: We are using 0HR_PY_PP_2 payroll posting documents data source which also extracts amount from PCL2 cluster table.

You might also face the lock issue as you are also trying to use the Payroll transaction data source.

Regards

KP

[former_member182516 · Accepted solution]

If you are still looking for any specific issues, then you can share so that we can provide solution/info for that issue.

[former_member182869 · Accepted solution]

Yes prashanth,

Is there any possibility for enabling other fields for inpopackage selections apart from PERNR and BW Month. since we want to extract data from Payroll Stndard Datasource, how delta mechanisam work for this extractor in the backend i need to do any extra settings for enabling delta and what is the frequency, i hope it is monthly once delta will run.

Regards
sree

[former_member182516 · Accepted solution]

Hari,

Yes you can have other fields in selection of IP, but ccould you let us know why you are looking for other fields in selections?

As part of standard HR data source other fields which are set to "No Selection Possible, Visibility Set" by SAP, if you want you can forcebily change the status by using simple program.

Delta upload: this is determine time range of evaluation for particular employee. Here delta will check

      - data which is already uploaded
      - not to upload the data of results where the period is not closed

If the above upload results of determined time range and evaluate voids then this will check for the time and date (time stamp) which is stored in the HR infotype 439.\

Data sources uses the FM : HRMS_BIW_EXTRACT_PY1

You can take help of ABAP person to debug and check how the delta is working for clear idea.

One main settings you need to take care is"Assignment of Payroll data to BW month" configuration settings.

Goto SBIW-> Settings for Application-specific Datasources(PI)-> Human Resources-> Payroll Plug-in-> Assignment of Payroll Data to BW month

You can go through the link provided by SUMAN for more clear idea on this configuration setting.

Regards

KP

[former_member182516]

Hi Hari,

Hope your system is already live n running with other data sources and all the configuration settings are in place.

Same setting will be applicable for the 0HR_PY_1 data source as well.

If you have enhanced the data source as per your business need then make sure the cmod code is written with best pratices so that the data extraction won't take much time.

Instead of loading all the historical data in a single shot you can split the load. first run the init without data transfe for the entire selection n later you can run full loads by spliting the 7 years data.

(0HR_PY_1 data source doesn't have any selection other than PERNR n  Calendar year / month. you need to split the data based on Calendar year / month and load the data.Its not possible to )

first give a try with one selection in PROD and see how much time a single data packet is taking to extract and the TRFC processing speed in SM58, if you feel the default data packet selection is taking time n TRFC are processing slow you can change the default transfer of data packets from source system in Info package from 50000 to 10000 n No of IDOcs to 3 or 4(by default it will be around 10).

RSCUSTV6 -- this will be set if its a fresh system if your system is already live no need to do any thing at this end.

Regards

KP

[RamanKorrapati]

Hi,

We can't guess coming problems.

As we know source have huge data, then better to load data with selections.

Try to load full loads with fiscal or fiscal period based

while loading data, try to play with data packet sizes and background parallel processes to speed up your loads.

Try to load this activity where there is no burden on ecc and bw. during your loads ask basis team to keep on eye on SM58. if any failures there, ask them to clear queues immediately.

Thanks