Quick tool to see if we can isolate some performance differences between different servers
The root cause seems to be the I/O for the nist-data-mirror doesn't use buffered input and is very slow. There's issue 115 and PR 116 open for these. It still showed a new server disk was configured very differently.
Just run the jar with the version of java you would like. For example:
java -jar java-hash-performance-0.0.1-SNAPSHOT.jar
mkdir ~/Downloads/nist/
cd ~/Downloads/nist/
java -jar ~/.m2/repository/us/springett/nist-data-mirror/1.5.4/nist-data-mirror-1.5.4.jar .