CH2++: A Hybrid Operational/Analytical Processing Benchmark for NoSQL Databases

History of CH2++

1. The initial CH benCHmark was proposed in 2011 to evaluate relational database systems with hybrid data management support

• CH borrows from and extends the schemas of TPC-C and TPC-H and combines the TPC-C transactions and the analytical queries of TPC-H to provide a hybrid benchmark.

2. CH2 was designed in 2021 as an extension of the CH benchmark to evaluate document database systems with hybrid data management support

• CH2 extends CH by introducing a document oriented schema, a data generation schema that creates a TPC-H like history, and modifying the CH analytical queries to be more along the lines of TPC-H.

3. CH2++ is a generalization of the CH2 schema to be more in line with JSON and analytics

• The TPC-C and TPC-H schemas have been modified to nest data where it would naturally be nested.
• CUSTOMER has been extended to have multiple addresses and phones and a set of categories.
• ITEM has been extended to have a set of categories.

---

Usage

Further documentation and all relevant source code for running the benchmark are located in the ch2driver/pytpcc directory.