On Ubuntu I'd install Elasticsearch using the repositories and then you can use the init script or systemd for the restart. If you use the zip or tar then you are responsible for killing elasticearch. You can just kill it.
I'm having the same issue with elasticsearch 2.3.3 on Debian 8. When asked for help (/usr/share/elasticsearch/bin/elasticsearch --help), the script claims that "elasticsearch start" is valid, but when I try it, I get that error message,
ERROR: Parameter [start]does not start with --
I like to be able to start elasticsearch like this, to be able to troubleshoot issues with the init script. On this node, the init script dies without logging anything.
You can run /usr/share/elasticsearch/bin/elasticsearch without anything and it should run Elasticsearch in the foreground. It isn't likely to have the right paths configured but it should start.
Usually when Elasticsearch doesn't start with the init script and doesn't log anything it is because you have a version of Java that can't load Elasticsearch - like 1.6. At least that is the first thing I check.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.