The Role of python in SAP HANA Database

[DINESH_S]

Dear All.

Why SAP uses Python in SAP HANA Database .The Role for python in SAP HANA Database.??

.

Reards,

Dinesh S.

[dvankempen]

Got the answer to your question? To improve the knowledge base and make the right answers easier to find, please mark the best answer and close the question. Thanks!

Don’t know how? See Ask and Answer Questions on SAP Community

For the readers, before you leave, don't forget to up/down vote the answers. You can vote on the question too.

Thanks.

[dvankempen · Accepted solution]

Hi Dinesh,

To elaborate on Jens' response;

Research for the project started back in 2006, as Dr. Hasso Plattner wrote in his paper A Common Database Approach for OLTP and OLAP Using an In-Memory Column Database from June 2009:

At the HPI, two and a half years ago, we started to analyze whether it is feasible or not to perform OLTP operations on an in-memory column store database. Many bachelor, master, and PhD projects focused on this topic.

As building blocks, three "databases" were used

• P*TIME - an in-memory row-store database SAP acquired in 2005
• MaxDB - another university research project this one going back to 1977 (!) acquired by SAP in 1998 and using this name since 2003 (with distribution partner MySQL for some time)
• TREX - the search engine used for NetWeaver (Text Retrieval and information EXtraction), going back to 1998.

In 2005, SAP released the SAP NetWeaver BI Accelerator BIA), which uses TREX to speed-up query performance of SAP BW (Business Warehouse), later renamed to BWA. BIA/BWA was made available as appliance using HPA (High-Performance Appliance) as internal reference (sounds familiar?). The nameserver, indexserver, and HDB script, for example (copyright 2003!) all originate here.

• https://blogs.sap.com/2017/11/28/sap-hana-under-the-hood-hdb-by-the-sap-hana-academy/

See also an earlier post about the origins of SAP HANA (the topic is also covered in my SAP HANA Introduction book by SAP Press).

• https://blogs.sap.com/2019/10/12/technology-overview-sap-hana-2.0-an-introduction/
• https://www.sap-press.com/sap-hana-20_4884/

Why Python?

Python 2.0 was just released in 2000 and it was the best technology at the time to write the regular expressions used by the early search engines, including Google (1997).Same for SAP and TREX.

• The Anatomy of a Large-Scale Hypertextual Web Search Engine | http://infolab.stanford.edu/~backrub/google.html

For the younger readers, once upon a time there was considerable choice in search engines: Altavista, Yahoo!, Excite, Lycos, Infoseek, Inktomi, ...

• https://www.thehistoryofseo.com/The-Industry/Short_History_of_Early_Search_Engines.aspx

[lbreddemann · Accepted solution]

Great answer!

I like to add the mySQL was a distribution partner - they didn’t contribute code to MaxDB, which was way further developed for enterprise use (ie recoverability) than MySQL back then. This was of course long before Oracle bought it.
For MaxDB it was a test to see if Open Source could work for this product; which it didn’t really work out.

thanks for this summary Denys!

[dvankempen · Accepted solution]

Thanks for the comment, Lars. Updated. Were you on the MaxDB team at the time?

[jgleichmann]

Hi,

just check the history of HANA. TREX based on python, BWA/BIA as well and some of the devolopers involved in TREX and BWA/BIA also developing HANA. So, the reason is historically.

Can you specify your question regarding the role? You can administrate the HANA with the python scripts. Some function of the HANA based on the scripts. For example the FSD (FullSystemDump), getParameter or topoInfo which are called by hdbnsutil.

Regards,

Jens