Learn how to choose the right SAP General Ledger extractor for your business needs among the different options provided by SAP.
Key Concept
The basic purpose of general ledger accounting is to provide a comprehensive view for external reporting. Until now, companies had to use multiple ledgers to meet reporting requirements for both external and internal management reporting. SAP General Ledger is the answer to get rid of multiple ledgers. It combines all the functions of multiple ledgers into a single ledger with multiple parallel ledgers.
With the introduction of the SAP General Ledger in SAP ERP Central Component (SAP ECC), companies are either migrating from their classic general ledger and special ledger to the new SAP General Ledger, or implementing it to support finance processes in SAP ECC. In either case, to support finance reporting out of SAP NetWeaver BW, you need to implement the new SAP General Ledger extractors offered by SAP. With the growing demand for real-time reporting within SAP NetWeaver BW, it is essential that you choose the appropriate extractors up front to avoid any surprises or rework. SAP has provided so many options that it can sometimes be confusing.
I’ll focus on the implementation considerations for the different SAP General Ledger extractors available, the prerequisites for implementing the new extractors, the extraction mechanism for the different extractors, the configuration steps for implementing these extractors, and common challenges and their resolution.
SAP General Ledger Extractor’s Technical Considerations
Figure 1 shows the integration of the different SAP ECC tables for SAP General Ledger and the information contained in those tables.
Figure 1
Extractor’s technical integration with SAP ECC tables
Table 1 compares balances and line item extractors.
Figure 1
Comparison of the SAP General Ledger balances and line item extractor
Note
Line item extractor 0FI_GL_14 is available as of SAP ECC 6.0 and enhancement package 3. However, the generated line item extractors 3FI_GL_XX_SI are available only after the activation of the business function FIN_GL_CI_1. Refer to SAP Note 1070629 for additional details.
Table 2 shows SAP General Ledger extractors for SAP NetWeaver BW for full versus delta extraction.
Figure 2
SAP General Ledger extractors
Configuration Steps for SAP General Ledger Extractors for SAP NetWeaver BW
Now I’ll show you the configuration steps for implementing the SAP General Ledger extractors for SAP NetWeaver BW. SAP has pre-delivered standard table groups in SAP ECC for activating the ledgers assigned to the FAGLFLEXT and FAGLFLEXA tables. The extract structure for these standard table groups is also delivered. To generate the SAP General Ledger extractors for the standard table groups, follow the steps below.
Step 1. Execute transaction code FAGLBW03 (Figure 2).
Figure 2
Generation of 3FI_GL_XX_TT extractors
Step 2. Double-click Totals Record-DataSource or Line Items DataSource to generate the extractor for the totals table (FAGLFLEXT) or line item table (FAGLFLEXA) (Figure 3), respectively. If you need to implement the balances extractor, choose the balances. Otherwise, you can choose the line items.
Figure 3
Generation of 3FI_GL_XX_SI extractors
Step 3. Choose the delta method (only applied to the 3FI_GL_XX_TT extractors). In Figure 2, I chose the radio button ADD.
Custom Table Group
You can create a custom table group in SAP ECC depending on specific requirements — for example, if you need to create a separate ledger in SAP ECC without affecting the data volume for the delivered standard FAGLFLEXA and FAGLFLEXT tables. SAP has not delivered the extract structure for the custom table group; therefore, you need to take an additional step to create the custom table structure before proceeding with the generation of the extractors. Follow the steps below to generate the extractors if you need a custom table group.
Step 1. Execute transaction code FAGLBW01 to generate the extract structure for the custom table group for the totals and line items table (Figure 4).
Figure 4
Generation of extract structure for custom table group
Note
When you activate the flag Totals Records check box under Generate Extract Structures for… in
Figure 4, the system generates an extract structure for the chosen totals record table. Enter the table group name in the space provided next to the totals table selection screen with the check box checked for either the Totals Records or Line Items. Then click the execute icon. The totals record extract structure receives the name of the totals table, and the T is replaced with a B.
Example: Table group GLFUNC:
Totals table > GLFUNCT
Totals record extract structure > GLFUNCB
When you activate the option to select extract structure Line Items, an extract structure is generated for the line items table belonging to the chosen totals record table. The line item extract structure receives the name of the totals table, and the T is replaced by a W.
Example: Table group GLFUNC:
Totals table > GLFUNCT
Line items extract structure > GLFUNCW
Step 2. Execute transaction code FAGLBW03.
Step 3. Double-click Totals Record-DataSource or Line Items DataSource to generate the extractor for the totals table (FAGLFLEXT) or line item table (FAGLFLEXA), respectively.
Step 4. Choose the delta method (only applied to the 3FI_GL_XX_TT extractors).
Which Is the Right Extractor for You?
Based on the general information provided so far about SAP General Ledger extractors, it may be confusing to choose the right extractor. Here are some key factors that can help you choose the right extractor for your scenario.
Factor 1: 0SALES vs. 0BALANCE Key Figure
The 0SALES key figure holds the period activities for a coding block combination (e.g., a coding block consists of fields such as company code, G/L account, profit center, or cost center). The 0BALANCE key figure holds the cumulative balance information for a coding block combination. It is essential that you choose the right key figure for your SAP NetWeaver BW/FI/CO implementation depending on your business needs. Depending on the delta type of the SAP NetWeaver BW extractor, the delivery of the key figure as part of the extractor can be tricky. See Table 3 to understand how SAP has delivered the key figures as part of different delta types.
Figure 3
Comparison of different key figures delivered by SAP General Ledger extractors
Here are some questions that you need to ask before choosing the key figure that is right for you.
Q1: Is there any reporting requirement for trending analysis across years? Or is there any outbound system (typically a planning/consolidation system) that requires complete trial balance information from the SAP BW system?
A1: For both requirements, the 0BALANCE key figure is the right option. However, the requirements can also be met by using 0SALES key figure with a lot of customization in the front end for meeting the trending requirements and subsequently a lot of customization with ABAP routines for the outbound requirements. This can lead to performance issues with huge volume constraints.
Q2: Is there any outbound interface that requires the complete trial balance dump from SAP NetWeaver BW every period?
A2: This can be satisfied via both 0SALES and 0BALANCE key figures. However, if data volume is a concern, 0BALANCE is the best bet.
Q3: Are there any concerns around reporting performance?
A3: With SAP NetWeaver BW Accelerator, both key figures give almost the same performance. Without SAP NetWeaver BW Accelerator, 0BALANCE is the best bet.
Factor 2: Delta Mechanism for SAP General Ledger Balances Extractor
One of the key aspects of the extractors is delta type. The delta type determines how the delta change information from SAP ECC is delivered to SAP NetWeaver BW. For example: for after image or additive image, SAP General Ledger extractors provide three types of deltas: AIED, ADD, and ADDD.
- AIED: This is the oldest of the SAP General Ledger balances extractors. Since it is an after image-based delta, a DataStore object (DSO) layer is required in overwrite mode in between the reporting layer. The extractor delta brings a lot of records and the performance is comparatively on the slower side. Refer to the factor 1 section for additional details. No additional ABAP code is needed for the 0BALANCE key figure.
- ADD: This works through a timestamp but goes directly to the SAP ECC-relevant tables and doesn’t read the delta queue directly. INIT reads from the FAGLFLEXT table and the delta reads from the FAGLFLEXA table along with BSEG, BSEG_ADD, and BKPF. You can load data to the reporting layer. No intermediate DSO is required. It extracts a smaller number of records from SAP ECC than ADDD, since it summarizes the entries on the SAP ECC side before sending it over to SAP NetWeaver BW. You need to write an ABAP routine in the transformation to populate the 0BALANCE key figures. Refer to the factor 1 section above for additional details.
- ADDD: This works through a timestamp-based delta queue. You can load data directly to the InfoCube. No intermediate DSO layer is required. Any changes to the balances in SAP ECC are captured in the delta queue through qRFC. You need to write an ABAP routine in the transformation to populate the 0BALANCE key figures. Refer to the factor 1 section for additional details.
Here are some questions that you need to ask before choosing the right delta mechanism option:
Q4: Which delta method is preferred for a huge data volume?
A4: The recommended delta type is ADD. Refer to the common challenges section at the end of this article for the SAP General Ledger balances extractor.
Q5: If you need a more frequent delta extraction for real-time reporting vs. standard OLAP reporting, which delta method should you use?
A5: The recommended delta type is ADDD. Refer to the common challenges section at the end of this article for the SAP General Ledger balances extractor.
Q6: If data volume is not a concern and real-time reporting is not needed, which delta method should you use?
A6: You can use the AIED method, but choose between ADD and ADDD. In this case, it doesn’t matter if you use ADD or ADDD as you do not need both data volume and real-time reporting. Q4 and Q5 represent the situation when you either go for ADD or ADDD.
Factor 3: More Frequent Data Extraction for Enabling Real-Time Reporting (RDA Factor)
There are often requirements to extract SAP General Ledger information from SAP ECC into SAP NetWeaver BW on a more frequent time interval than once or twice a day. If a more frequent extract is required, SAP has proposed a Real-time Data Acquisition (RDA)-based daemon update to meet the requirements. However, it may not be the best suited option for your needs as it comes with certain limitations. I’ll compare the RDA method with an alternative:
- RDA-based daemon update: RDA supports tactical decision-making. It also supports operational reporting by allowing you to send data to the delta queue or a Persistent Staging Area (PSA) table in real time. You then use a daemon to transfer DSOs to the DSO layer at frequent regular intervals. The data is stored persistently in SAP NetWeaver BW. SAP recommends that you use RDA if you want to transfer data to SAP NetWeaver BW at shorter intervals (every minute) than scheduled data transfers and you need up-to-date data to be regularly available in reporting (several times a day, at least).
However, the major drawback in this approach is that you can only use RDA to fill DSOs. A two-step data transfer is supported. Data is first transferred into the PSA and then into the DSO. The DSO cannot be used as the source for a further real-time data transfer to another DSO.
- An alternative approach to RDA: Refer to factor 2 (different delta methods). Since the ADDD method updates the delta queue directly at the time of posting through qRFCs, the delta queue is already pre-populated for extraction. Schedule an InfoPackage to pull data from the delta queue a couple of minutes into PSA and a Data Transfer Process (DTP) to push data directly to the reporting layer InfoCubes. The InfoPackage and DTP can be combined into a process chain. You can recursively call the process chain to extract data and push it to the reporting layer InfoCubes as soon as the last delta load is completed. You have more flexibility with this option compared to the RDA-based daemon update for real-time reporting.
Q7: Is RDA the only option for real-time reporting?
A7: Yes and no. This is the standard option that is offered by SAP but it comes with a lot of limitations and drawbacks. Refer to the SAP Help link above for the details on the disadvantage of using the RDA-based daemon update. If data volume is a concern, real-time reporting through RDA is not an option. My recommendation is to explore the alternative approach mentioned above in all cases (i.e., huge data volume as well as small data volume). Also, carefully review the limitations of the RDA approach before choosing this option.
Which Is the Right Extractor for You?
To choose the right extractor for your needs, keep in mind the following factors.
Factor 1: Leading Ledger vs. Non-Leading Ledger Data
To extract SAP General Ledger line-item information for the leading ledger, either use 0FI_GL_14 (available as of enhancement package 3 for SAP ECC 6.0) or generated extractor 3FI_GL_XX_SI (where XX represents the leading ledger). This extractor is available as of enhancement package 3 for SAP ECC 6.0 and business function FIN_GL_CI_1).
Factor 2: SAP General Ledger Flag Activation Contention with Business Function Activation
If you are migrating from the classic general ledger and special purpose ledger to the SAP General Ledger, SAP General Ledger flag activation is a key step in the activation. If a multi-weekend go-live approach is followed at your company, carefully think through the strategy for SAP General Ledger line-item extractors. A multi-weekend go-live approach means you go live over multiple weekends with some functionality going live earlier than others. The normal production activities are not interrupted during the week.
Q8: Is there any option to extract general ledger line-item data into SAP NetWeaver BW prior to SAP General Ledger flag activation on the SAP ECC side, but after data migration into the FLEX* tables (using migration cockpit)?
A8: Yes. You can use extractor 0FI_GL_14, which brings only leading ledger data. Refer to SAP Note 1070629 for additional details.
Factor 3: Data Volume and Performance
Performance is a concern for the general ledger line-item extractors in general. However, there is not much difference between the extractors when it comes to performance. I have provided some tips and tricks in the common challenges section.
After going through all the different factors, what does it mean to you? Based on my experience going through a full-blown implementation, I have provided information about the extractors that is not documented or that is hard to find at one place that may help you make the right decision for your company and avoid costly mistakes.
To further simplify the discussion, follow through the decision tree shown in Figure 5 to help you find the right SAP General Ledger extractor for your upcoming finance implementation.
Figure 5
SAP General Ledger extractors for SAP NetWeaver BW — decision tree
Common Challenges: Balances Extractor
You face a few common challenges when you implement SAP General Ledger extractors. These are not documented in one place by SAP, making it very difficult to gather the information prior to the implementation, and companies realize it after the fact. I have gathered and consolidated some of the known challenges for the SAP General Ledger balances extractor so that you can make an informed decision.
Common Challenges — SAP General Ledger Line-Item Extractor
Here are some common challenges you may face while implementing SAP General Ledger line-item extractors.
- The INIT without data transfer concept doesn’t work with SAP General Ledger line-item extractors. Choose to use INIT with the data transfer option.
- INIT with data transfer takes a lot of time for a given year and period. A given year and period combination might take a lot of time depending on the data volume in the FAGLFLEXA table on the SAP ECC side. However, use the fields RYEAR and POPER for initialization instead of GJAHR to improve the performance considerably.
- Opening balances are not extracted by the SAP General Ledger line-item extractors
- Plan data is not extracted by the SAP General Ledger line-item extractors
- Performance of the SAP General Ledger line-item extractor is poor in general. To improve the performance of the extractor, follow these steps as the first resort:
- Step 1. Create a secondary index on the TIMESTAMP field in FAGLFLEXA table on the SAP ECC side
- Step 2. Run the RUNSTATS on a daily basis on this table if there is a 10 percent change in overall data volume
- Step 3. Increase the size of the data packets transferred between SAP ECC and SAP NetWeaver BW
After following these steps, if the performance benefits are not realized, evaluate the SAP Notes in Table 4 to see what applies to your scenario.
Table 4
SAP Notes to improve performance
Soumya Mishra
Soumya Mishra is a senior BI practitioner with a total of eight years of experience in the SAP BI space. He has been continuously focused on developing business enabling analytic strategies and solutions. He has extensive enterprise analytics assessment knowledge and has successfully played a key role in delivering large-scale finance transformation solutions that deliver actionable information and insight to users, giving them the ability to make informed decisions. He has deep knowledge of BI, and SAP ECC FI/CO configuration, as well as experience with implementations in several Fortune 500 companies in the consumer products and financial services industries. He is a certified SAP BI solution consultant with deep experience in development and integration of BI strategy; process redesign with information systems; and project planning and implementations with strong communication, organizational, and project management skills. You can view his LinkedIn profile at https://www.linkedin.com/profile/view?id=50199012&locale=en_US&trk=tab_pro.
You may contact the author at srmishra@us.ibm.com.
If you have comments about this article or publication, or would like to submit an article idea, please contact the editor.
