Get bulk data out of SAP: the performance comparison of SAP ETL solutions

Background

SAP is complex. The sophisticated yet powerful data models in SAP is what makes SAP the most successful business software in the world. Yet the complexity makes it very challenging for the majority of the ETL tools on the market to extract bulk data from SAP. Even for the tools which claim to be able to extract from SAP, a lot of those rely on SAP's native read table function to fetch limited amount of data from SAP, specifically with limitations on the number of rows and columns. Ignoring those immature solutions, here we will look at the performance comparisons among SSIS+AecorSoft Integration Service, SAP's BOBJ Data Services, and SSIS+another third-party SSIS custom component.

Test Setup

SAP system: SAP IDES Ehp7 running on quad core I7, 16GB RAM
ETL platforms: all running on Windows 2012R2, quad core i7, 8GB RAM
ETL job setup: simple workflow with SAP data source and File destination to local drive

 

Test1: General Ledger Header BKPF table

# of rows: 584,448
# of columns: 99
rough size on disk: 200MB or so.

Solution


SSIS 2016 + AecorSoft Integration Service

Time Taken


24 seconds

Detail



SSIS 2016 + third party

57 seconds


BOBJ Data Services 4.2

83.8 seconds


 

Test2: General Ledger Line Item extractor 0FI_GL_4

# of rows: 1,678,342
# of columns: 76
rough size on disk: 1.2GB or so

Solution


SSIS 2016 + AecorSoft Integration Service

Time Taken


344 seconds

Detail



SSIS 2016 + third party

524 seconds


BOBJ Data Services 4.2

420 seconds

BODS 0FI_GL_4 perf.PNG

 

Conclusion

Here we can see how much faster AecorSoft Integration Service is at extracting bulk SAP data, AecorSoft is not only much more performing than other comparable products, but also much easier to config with intuitive interfaces, while in full compliance of SAP application layer security. The highly efficient and intelligent algorithms are the core of AecorSoft product suite.