jubilee vs torqbox performance on techempower json benchmark after warmup period

jubilee

# Jubille started with:
#    jubilee -p 8080 -e production
#
# Note that the Jubilee console outputs quite a few errors like below during
# the test:
#
# Jan 03, 2014 2:15:31 PM org.vertx.java.core.logging.impl.JULLogDelegate error
# SEVERE: Unhandled exception
# java.nio.channels.ClosedChannelException


$ wrk -H 'Host: localhost' -H 'Accept: text/plain,text/html;q=0.9,application/xhtml+xml;q=0.9,application/xml;q=0.8,*/*;q=0.7' -H 'Connection: keep-alive' -d 15 -c 256 -t 8 http://localhost:8080/json
Running 15s test @ http://localhost:8080/json
  8 threads and 256 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency    13.51ms    1.53ms  35.08ms   81.11%
    Req/Sec     2.40k   412.76     4.45k    69.98%
  280662 requests in 15.00s, 25.96MB read
Requests/sec:  18710.60
Transfer/sec:      1.73MB

torqbox

# TorqBox started with:
#     torquebox -p 8080 -e production

$ wrk -H 'Host: localhost' -H 'Accept: text/plain,text/html;q=0.9,application/xhtml+xml;q=0.9,application/xml;q=0.8,*/*;q=0.7' -H 'Connection: keep-alive' -d 15 -c 256 -t 8 http://localhost:8080/json
Running 15s test @ http://localhost:8080/json
  8 threads and 256 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency     2.11ms    2.12ms 221.28ms   93.13%
    Req/Sec    15.67k     2.54k   26.67k    69.53%
  1772646 requests in 15.00s, 297.53MB read
Requests/sec: 118184.53
Transfer/sec:     19.84MB

Those exceptions are because wrk cut off the connections at the end of the test. I'll see if they can be caught in my end or it has to be done in vertx. Also did you warm the JVM up before running the benchmarks? (Run the benchmark several iterations before get the final numbers, so that method hotspots can get inlined and JITed).

Did we get any more accurate information here @isaiah?