OodJob

Library that provides a generic interface to submit/status/hold/release/delete batch jobs for various resource managers.

Installation

Add this line to your application's Gemfile:

gem 'ood_job'

And then execute:

$ bundle

Or install it yourself as:

$ gem install ood_job

Usage

Create a Job Script

First we will need a script to submit to the resource manager. The simplest script object consists of just a string of shell code.

# Define a simple script
script = OodJob::Script.new(content: "echo 'hello world'")

With time you will create more complicated script objects:

script = OodJob::Script.new(
  content: Pathname.new("/path/to/my/script").read,
  job_name: "my_solver_job",
  wall_time: 3600,                         # walltime=01:00:00
  nodes: [NodeRequest.new(procs: 12)]*20,  # nodes=20:ppn=12
  output_path: Pathname.new("/path/to/output"),
  join_files: true,
  accounting_id: "PZS0001"
)

Note: Any options supplied when creating a script object will override those specified in the script file itself.

Submit a Job

Once you have a script object you are satisfied with, you will want to submit it to a given resource manager. This is where the OodCluster::Cluster object (ood_cluster) will come into play. This object describes a given cluster and its various servers (one of which is a resource manager).

If you are using the OodAppkit Rails engine (ood_appkit) from a Rails app:

# Get the Oakley cluster object from OodAppkit
oak = OodAppkit.clusters['oakley']

# Use this cluster object to submit the script (be sure you choose the correct
# adapter to communicate with this cluster)
adapter = OodJob::Adapters::Torque.new(cluster: oak)

# Submit your script
adapter.submit(script: script)
#=> "1234.server"

To submit dependencies for complex workflows:

# Submit pre-processing script
pre_id = adapter.submit(script: pre_process_script)

# Submit multiple solver scripts to run if pre-processing terminated
# successfully
solve_id1 = adapter.submit(script: solver_script1, afterok: [pre_id])
solve_id2 = adapter.submit(script: solver_script2, afterok: [pre_id])
solve_id3 = adapter.submit(script: solver_script3, afterok: [pre_id])
solve_id4 = adapter.submit(script: solver_script4, afterok: [pre_id])

# Sumit post-processing script to run after all jobs terminated with or without
# errors
adapter.submit(script: post_process, afterany: [solve_id1, solve_id2, solve_id3, solve_id4])

Status of Job

Assuming we have our adapter object from before and we know the job id:

# Get job status
status = adapter.status(id: "1234.server")
#=> #<OodJob::Status @state=:running>

# Check if job is queued or running
status.queued?  #=> false
status.running? #=> true

If we want more details about the job, we can request its #info

# Get job info
info = adapter.info(id: "1234.server")
#=> #<OodJob::Info ...>

# Check some values
info.job_owner       #=> "bob"
info.submission_time #=> 2016-11-07 10:12:05 -0500
info.procs           #=> 12
info.status          #=> #<OodJob::Status @state=:running>
...

Note: You can retrieve the info for every job on the cluster if you specify a blank id.

# Get all jobs info
adapter.info(id: "")
#=> [
#     #<OodJob::Info ...>,
#     #<OodJob::Info ...>,
#     ...
#   ]

Hold a Job

Assuming we have our adapter object from before and we know the job id:

# Hold job
adapter.hold(id: "1234.server")

Release a Job

Assuming we have our adapter object from before and we know the job id:

# Release job
adapter.release(id: "1234.server")

Delete a job

Assuming we have our adapter object from before and we know the job id:

# Delete job
adapter.delete(id: "1234.server")

Development

To develop you will need to install all the necessary gems:

bundle install --path=vendor/bundle

You can then access the development console through:

bundle exec pry -Ilib

You must always run through the tests after changes to make sure you didn't break anything:

bundle exec rspec spec

Contributing

  1. Fork it ( https://github.com/[my-github-username]/ood_job/fork )
  2. Create your feature branch (git checkout -b my-new-feature)
  3. Commit your changes (git commit -am 'Add some feature')
  4. Push to the branch (git push origin my-new-feature)
  5. Create a new Pull Request