The Echoprint server can insert and query for millions of tracks to match a song given an Echoprint code. It is based on Apache Solr and also uses Tokyo Tyrant, a fast key-value store. 

Echoprint’s server code (the Python glue and matching code as well as the Solr extensions) is available under the Apache 2.0 License. Solr and Tokyo Tyrant are licensed separately (Apache and LGPL respectively.) The server also includes bigeval, our software suite for evaluating fingerprint accuracy.

Note that in normal use you do not need to boot an Echoprint server. You can use Echo Nest song/identify, who has an Echoprint server already booted. Over time others will also boot servers hosting a mirror of the Echoprint data.

Visit the echoprint-server GitHub page for the source and installation instructions.