Skip to Content.
Sympa Menu

opal - Re: [Opal] Docker build

opal AT lists.psi.ch

Subject: The OPAL Discussion Forum

List archive

Re: [Opal] Docker build


Chronological Thread 
  • From: "Adelmann Andreas (PSI)" <andreas.adelmann AT psi.ch>
  • To: Rob Nagler <opal-rtnkh AT q33.us>, "bruhwiler AT radiasoft.net" <bruhwiler AT radiasoft.net>
  • Cc: opal <opal AT lists.psi.ch>, Eric Prebys <eprebys AT ucdavis.edu>
  • Subject: Re: [Opal] Docker build
  • Date: Thu, 15 Aug 2019 23:39:30 +0000
  • Accept-language: en-US, de-CH
  • Authentication-results: localhost; iprev=pass (edge20.ethz.ch) smtp.remote-ip=82.130.99.26; spf=pass smtp.mailfrom=psi.ch; dmarc=skipped

Hi Rob and David, awesome work and very much needed (and important)  for reproducible science! 

Cheers Andreas 
------
Dr. sc. math. Andreas (Andy) Adelmann
Head a.i. Labor for Scientific Computing and Modelling 
Paul Scherrer Institut OHSA/ CH-5232 Villigen PSI
Phone Office: xx41 56 310 42 33 Fax: xx41 56 310 31 91
Phone Home: xx41 62 891 91 44
-------------------------------------------------------
Friday: ETH HPK G 28   +41 44 633 3076
============================================
The more exotic, the more abstract the knowledge, 
the more profound will be its consequences.
Leon Lederman 
============================================

On 16 Aug 2019, at 00:00, Rob Nagler <opal-rtnkh AT q33.us> wrote:

Hi Eric,

As David noted, we build a Docker image "radiasoft/beamsim" with OPAL and many other codes included. It currently contains OPAL 2.01.

One way to run this image on Mac or Linux is:

mkdir beamsim
cd beamsim
curl radia.run | bash

Another way is to run it in the current directory:

curl radia.run | bash -s beamsim

This does a few things that you'll want, e.g. set the run user and mount the current directory. 

If you are running Windows or are familiar with Docker, you can also run manually with:

docker run -it -u vagrant radiasoft/beamsim:prod bash -l

Another option is to run a local Jupyter server with:

curl jupyter.run | bash

This will download the image radiasoft/beamsim-jupyter, and invoke it with the appropriate arguments as well as tell you how to connect to the server.

Note that these images are very large (10GB), because they contain a lot of codes, which are integrated into a single pyenv for seamless execution of codes with Python 2 or 3. Expect a long "docker pull" even when you are on a fast network (most of the time is spent in the "extract" phase).

It would be helpful to know your use case, because we have support for efficient Docker image composition. We could potentially build a smaller Docker image that would better suit your needs. If you are interested in discussing further, please email me off list.

Cheers,
Rob





Archive powered by MHonArc 2.6.19.

Top of Page