Github user MrSandmanRUS commented on a diff in the pull request: https://github.com/apache/phoenix/pull/311#discussion_r203552153 --- Diff: bin/traceserver.py --- @@ -116,8 +116,10 @@ # " -Xdebug -Xrunjdwp:transport=dt_socket,address=5005,server=y,suspend=n " + \ # " -XX:+UnlockCommercialFeatures -XX:+FlightRecorder -XX:FlightRecorderOptions=defaultrecording=true,dumponexit=true" + \ -java_cmd = '%(java)s $PHOENIX_OPTS ' + \ - '-cp ' + hbase_config_path + os.pathsep + phoenix_utils.phoenix_traceserver_jar + os.pathsep + phoenix_utils.phoenix_client_jar + \ +java_cmd = '%(java)s ' + \ --- End diff -- https://issues.apache.org/jira/browse/PHOENIX-2659 I think it's your comment: "Josh Elser added a comment - 12/Feb/16 03:23 ok, PHOENIX_OPTS isn't working because subprocess isn't popping a shell like sqlline does. It will automatically get added to PQS's environment already. We can just remove that. Steve is spot-on for the argument parsing. It's just busted." So, I just delete it.
---