Unum Cloud Serving Benchmark is the grandchild of Yahoo Cloud Serving Benchmark, reimplemented in C++, with less costly abstractions and with more workloads, crafted specifically for the Big Data age! A full description of this benchmark can be found here, in our corporate blog, together with the recent results.
To run the benchmark, call ./run.py.
Output will be placed into the bench/results/ folder.
- WiredTiger. Version 10.0.0.
- LevelDB. Version 1.23.
- RocksDB. Version 6.29.3.
- LMDB. Version 0.9.29.
Our UnumDB is also evaluated with this benchmark, but it wasn't included into the public mirror.
- ∅: imports monotonically increasing keys 🔄
- A: 50% reads + 50% updates, all random
- C: reads, all random
- D: 95% reads + 5% inserts, all random
- E: range scan 🔄
- ✗: batch read 🆕
- Y: batch insert 🆕
- Z: scans 🆕
The ∅ was previously implemented as one-by-one inserts, but some KVS support the external construction of its internal representation files. The E was previously mixed with 5% insertions.
- Current benchmarks don't use custom key comparators. Both variants were tested and it didn't affect the speed.
- WiredTiger sometimes crashes on 1 TB benchmarks.
- Read/Update might be replaced with a Read-Modify-Write operation.