Open kevin-klein opened 8 years ago
The Java Fact Extractor has very poor startup time compared to other fact extractors:
kevin@kevin-Aspire-E1-572G:~/101worker/extractors/Java$ time echo 'class Java { }' | ./extractor { "package": "", "imports": [], "fragments": [ { "classifier": "class", "name": "Java", "fragments": [], "annotations": [], "startLine": 1, "endLine": 1 } ] } real 0m0.214s user 0m0.300s sys 0m0.020s
Javascript for comparison
kevin@kevin-Aspire-E1-572G:~/101worker/extractors/JavaScript$ time echo 'function x() { console.log("hello") };' | ./extractor [ { "classifier": "function", "name": "x", "fragments": [] }, { "type": "EmptyStatement" } ] real 0m0.064s user 0m0.060s sys 0m0.004s
The JVM used is openjdk:
kevin@kevin-Aspire-E1-572G:~/101worker/extractors/JavaScript$ java -version openjdk version "1.8.0_66-internal" OpenJDK Runtime Environment (build 1.8.0_66-internal-b17) OpenJDK 64-Bit Server VM (build 25.66-b17, mixed mode)
Switching the JVM mode to client did not improve the performance, in fact it made a little worse.
The Java Fact Extractor has very poor startup time compared to other fact extractors:
Javascript for comparison
The JVM used is openjdk:
Switching the JVM mode to client did not improve the performance, in fact it made a little worse.