Plow is render farm management software specifically designed for VFX workloads.
There are currently no official releases of Plow, so the following instructions will get you started with a source checkout of Plow. Please report any issues either directly on github, or in the Plow Users Google forums: http://groups.google.com/group/plow-render
- Postgresql 9.2
- Java7
Plow has an API for job submission, but the Blueprint library makes this much easier.
The plow server acts as the central brain for your render farm. It contains the plow dispatcher and exposes a thrift API for interacting with jobs.
The plow server code compiles to a Web Application Archive file, or WAR file. This file must be hosted by within a Java application server.
While developing Plow, you will also need to compile your own WAR file at times.
For some reason this version isn't in Maven yet. This will install the Postgres JDBC driver into your local Maven repository.
$ cd server/lib
$ ./pginstall.sh
Next step, must build the Plow thrift bindings for Java and install those in your local Maven repo.
$ cd lib/thrift
$ ./generate-sources.sh
$ cd ../java
$ mvn install
Now you can actually build the server.
$ cd server
$ mvn package
This will create a file named target/plow.war. Replace the plow.war file in the executable server package you downloaded and your now on the latest version. You can also copy (or symlink) plow.war into dist/webapps and run start it using dist/start-plow.sh:
$ cp target/plow.war dist/webapps/
OR
$ ln -s target/plow.war dist/webapps/
If the thrift files change at all, you have to re-generate + install the thrift bindings and recompile the server.
Install Postgresql 9.2.
Create a database called 'plow' for user 'plow'. set password: plow (This is configurable in the plow-server-bin/resources/plow.properties )
Execute the sql file:
$ psql -h <hostname> -U <username> -d <dbname> -f server/ddl/plow-schema.sql
The latest Python client can be install from the source checkout using the following:
(first make sure to generate the thift bindings)
> cd lib/python
> python setup.py install
This will attempt to also copy the etc/plow/*.cfg files to /usr/local/etc/plow/. You may also manually copy them to
~/.plow/ instead if you want.
For Development...
> cd lib/python
> python setup.py develop
This will build the source files in place, and put a linked egg in your PYTHONPATH.
You can also set the PYTHONPATH path environment variable for development.
> cd <plow source directory>/server/dist/
> ./start-plow.sh
If Java7 is not in your path, plow will pick it up if the JAVA_HOME env var is set. On Mac, this will
be something like this:
> export JAVA_HOME="/Library/Java/JavaVirtualMachines/jdk1.7.0_10.jdk/Contents/Home"
> ./start-plow.sh
First thing you need to do if you are using a git checkout of plow is setup your environment.
$ export PLOW_ROOT="/path/to/plow/checkout"
$ export PYTHONPATH="/path/to/plow/checkout/lib/python"
$ export PATH="$PATH:$PLOW_ROOT/bin"
OR if you are in bash, you can just source setup_env.sh located on the root level. This should set all your environment variables, including all config env variables.
$ source <plow source directory>/setup_env.sh
Currently supported on Mac/Linux
If you have installed the client tools using the setup.py, then you should now have rndaemon command in your path:
$ rndaemon
Otherwise, you can launch rndaemon from your git checkout (after setting your environment variables and ensuring you have all python dependencies installed).
$ <plow source directory>/bin/rndaemon
The daemon will first look for an optional config file explicitely set with the PLOW_RNDAEMON_CFG environment variable:
$ export PLOW_RNDAEMON_CFG="/path/to/etc/plow/rndaemon.cfg"
Otherwise, it will search for: /usr/local/etc/rndaemon.cfg, $PLOW_ROOT/etc/plow/rndaemon.cfg, and then ~/.plow/rndaemon.cfg
After installing the python client, and having PySide installed, you should be able to run:
$ plow-wrangler