What is KingDB?
KingDB is a fast on-disk persistent key-value store. You can embed it or use it as a library in your C++ applications.
KingServer is a server application that embeds KingDB and implements the Memcached protocol. It allows you to access your data through a network interface using whatever programming language you want. KingServer is not a distributed system: the data lives in a single machine.
WARNING: This is still alpha code. Even though unit-tests are covering the core components, make sure you run tests for your environment before using KingDB in production.
Why use KingDB?
- Fast for heavy write workloads and random reads.
- The architecture, code, and data format are simple.
- Multipart API to read and write large entries in smaller parts.
- Multiple threads can access the same database safely.
- Crash-proof: nothing ever gets overwritten.
- Iterators and read-only consistent snapshots.
- Compaction happens in a background thread, and does not block reads or writes.
- The data format allows hot backups to be made.
- Covered by unit tests.
How fast is KingDB?
KingDB was benchmarked using the same test suite as LevelDB. On a Linux CentOS 6.5, for entries with 16-byte keys and 100-byte values (50 bytes after compression), the performance was:
|Workload||Operations per second|
For more details and a comparison with LevelDB, you can refer to the full KingDB benchmarks.
Where is the documentation?
How to install KingDB?
You can find installation instructions in the installation section of the documentation.
KingDB has no external dependencies and has been tested on:
- Mac OS X 10.9.5 with Apple LLVM version 6.0 (clang-600.0.51)
- Linux Ubuntu 14.04 x64 with GCC 4.9.2
- Linux Ubuntu 15.04 x64 with GCC 4.9.2-10ubuntu13
- Linux CentOS 6.5 x86_64 with GCC 4.9.2
If you are using GCC, update the Makefile and add -fno-builtin-memcmp in the CFLAGS, and if you have tcmalloc on your system, add -ltcmalloc to the LDFLAGS. This will give you a nice performance speed-up.