I heard a buzzword "Optimistic Latch-Free Index Traversal" in one blog post on Introduction to SAP HANA by Dr. Vishal Sikka. I googled about it in context of SAP HANA Architecture but didn't find anything concrete. Could someone please explain this whole concept thoroughly, in context of SAP HANA?
This refers to a core part of the former PTime* in-memory rowstore DBMS, which served as a technology foundation for some parts of SAP HANA, e.g. some internal row store structures (CSB+ tree index types).
These index type memory structures are used internally by SAP HANA to organize and process data in memory with many parallel cores.
A little bit of googling would have led you to the paper that introduced the CSB+ trees:
Cache-Conscious Concurrency Control of Main-Memory Indexes on Shared-Memory Multiprocessor Systems
Given that the rowstore in SAP HANA becomes less and less important this probably has more value if you want to build your own in-memory row-store database.
To me, knowing this paper made zero difference to my abilities to use SAP HANA.
Add a comment