I just pushed Java Service Wrapper support for elasticsearch. The idea of
Java Service Wrapper is to have a lightweight process forking the actual
Java process that is elasticsearch. This process allows to run elasticsearch
in console mode, as well as to start it in the background, and stop it
(finally, a command line option to stop it). The script also allows to
install elasticsearch in init.d/launchd(mac)/windows service. The scripts
exists under the bin/service directory. On top of that, the shutdown API now
automatically uses service wrapper when its run under it to send the stop
command through the service wrapper.
Also, a restart API as been added (similar to the shutdown API, just
replace shutdown with restart). When running in regular mode, it will
restart elasticsearch within the same JVM. When run under the service
wrapper, the wrapper will restart the whole JVM and starting a fresh JVM
with a fresh elasticsearch node.
The wrapper also provides a nice ping ability to detect if the process is
hung, and if so, it will restart it automatically. The default ping timeout
is 5 minutes, but all is configurable under the
bin/service/elasticsearch.conf.
A note regarding the service wrapper. Memory configuration (min/max) are
configured within the elasticsearch.conf file. All the rest of elasticsearch
configuration can be configured in any case through the
elasticsearch.(yml|json) files.
Hi,
I just pushed Java Service Wrapper support for elasticsearch. The idea of
Java Service Wrapper is to have a lightweight process forking the actual
Java process that is elasticsearch. This process allows to run elasticsearch
in console mode, as well as to start it in the background, and stop it
(finally, a command line option to stop it). The script also allows to
install elasticsearch in init.d/launchd(mac)/windows service. The scripts
exists under the bin/service directory. On top of that, the shutdown API now
automatically uses service wrapper when its run under it to send the stop
command through the service wrapper.
Also, a restart API as been added (similar to the shutdown API, just
replace shutdown with restart). When running in regular mode, it will
restart elasticsearch within the same JVM. When run under the service
wrapper, the wrapper will restart the whole JVM and starting a fresh JVM
with a fresh elasticsearch node.
The wrapper also provides a nice ping ability to detect if the process is
hung, and if so, it will restart it automatically. The default ping timeout
is 5 minutes, but all is configurable under the
bin/service/elasticsearch.conf.
A note regarding the service wrapper. Memory configuration (min/max) are
configured within the elasticsearch.conf file. All the rest of elasticsearch
configuration can be configured in any case through the
elasticsearch.(yml|json) files.
cheers,
shay.banon
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.
