Usage

This document explains the basic usage of PenchY.

There is only a single entry point to PenchY, the penchy command, which takes quite a few command line parameters:

• -h or --help shows a help message and exit

• -c CONFIG or --config config specifies a configuration file. Please consult the Configuration file section in order to learn how to write such a file. This argument defaults to ~/.penchyrc.

• --logfile specifies the file to which PenchY will log. Please note that PenchY will automatically rotate logfiles, so if you pass --logfile foo.log, foo.log.1 might be created by PenchY.

• -d or --debug sets the log level to DEBUG.

• -q or --quiet sets the log level to WARNING.

• --check checks a job for validity only.

• --visualize visualize the dependencies of the job’s pipelines as Graph (needs Graphviz).

• --run-locally runs a job locally without the involvement of client/server communication. This requires the hostname passed to NodeSetting to be localhost.

• -f or --load-from will load PenchY from the path supplied instead of acquiring it using maven. This is a pretty nifty feature if you are working on the client code and don’t want to deploy the changes using maven as you write it. A simple scenario might be copying your PenchY version to the nodes using rsync and then using that version. Let’s say you are using rsync to copy your version of PenchY to /tmp/penchy2, then you’d need to pass -f /tmp/penchy2 to the penchy command on the server, which will result in all nodes loading PenchY from this path. The following snippet might be useful to you:

  #!/bin/bash -e
  
  for node in 192.168.56.10 192.168.56.11; do
      rsync -az --exclude '*.log' --exclude '*.pyc' . bench@${node}:~/penchy
  done
  
  bin/penchy --load-from /home/bench/penchy $*