Welcome!
How does this work?
What you're looking at right now is Workbench, the graphical interface to the Arvados system.
For example, here's what a pipeline running in Arvados looks like:
<%= image_tag "pipeline-running.gif", :class => "style_image" %>
Click the Next > button below for a speed tour of Arvados.
Note: You can always come back to this Getting Started guide by clicking the in the upper-right corner.
Quickstart
Don't like reading manuals? Get started by running your first pipeline in 3 quick steps:
Tip: log-in or register with any google account if you haven't already
- Go to the Dashboard > Run a pipeline...
<%= image_tag "mouse-move.gif", :class => "style_image" %>
- Mason Lab -- Ancestry Mapper (public) > Next: choose inputs
- Run
Voila!
Your pipeline is now spooling up and getting ready to run!
Go ahead, try it for yourself right now.
Or click
Next > below to keep reading!
Three Useful Terms
* Pipeline -- A re-usable series of analysis steps.
** Sometimes known as a “workflow” in other systems
** A list of well-documented public pipelines can be found in the upper right corner by clicking the "?" > "Public Pipelines and Datasets"
** Pro-tip: Pipeline > Jobs > Tasks. A pipeline contains jobs which contain tasks.
** Pipelines can only be shared within a project
* Collection -- Like a folder, but better
** Upload data right in your browser
** Better than a folder?
*** Collections contain the content-address of the data instead of the data itself
*** Sets of data can be flexibly defined and re-defined without duplicating data
** Collections can be shared using the "Sharing and Permissions" > "Share" button
* Projects -- Contain pipelines templates, pipeline instances (individual runs of a pipeline), and collections
** The most useful one is your default "Home" project, under Projects > Home
** Projects can be shared using the "sharing" tab
1. Reproducible Analyses: Enough said.
2. Data provenance: Every file in Arvados knows can tell you where it came from.
3. Serious scaling: Need 500 GB of space? 200 compute hours? Arvados scales and parallelizes your work for you intelligently.
4. Share pipelines or data: Easily publish your work the world, just like the Pathomap team did: http://www.pathomap.org/2015/04/08/run-the-pathomap-human-ancestry-pipeline-on-arvados/
5. Use existing pipelines: Use best-practices pipelines on your own data with the click of a button
6. Open-source: Arvados is completely open-source. Check us out at http://arvados.org/
This guide and even all of Workbench is just a glimpse into the power of Arvados. Want to use the command-line instead? Or hungry to learn more? Check out our detailed documentation: http://doc.arvados.org/ (our real-time contact info is there too!)
That's all, folks!