phoebe is a fast read only, in memory, key-value database written in C++

• uses kafka compacted topics for persistense.
• a typical use-case would be a authentication database for users in the range of 10M users per server

there are also some useful tools that seems to handle what hadoop/hive can generate.

• avrofile2kafka: inserts avrofiles in kafka using confluent.io schema registry

• avrofile2postgres: inserts avrofiles in a postgres db

Building

• first see https://github.com/bitbouncer/csi-build-scripts
• bash build_linux.sh (should work on centos7 and ubuntu14)
• rebuild_win64_vc12.bat (windows)