Loci-Stream on AWS

Loci-Stream is now available to run on Amazon Web Services (AWS). This option is good for customers who may only have a short-term need for simulation capabilities, or for customers that do not have dedicated on-premises hardware. With Loci-Stream on AWS, users can quickly allocate computer hardware that is needed for performing a simulation, run the simulation on this hardware, and then download the generated data from the AWS servers to local machines for analysis. Users are charged an hourly-based software fee for running Loci-Stream, as well as an hourly-based hardware fee that is dependent on the type of hardware that is selected. Potential customers who may want to assess the capabilities of Loci-Stream can also utilize the AWS framework to run our tutorial cases for a relatively small fee.

AWS Overview

The AWS framework is quite vast and can be intimidating at first glance. Amazon has very extensive documentation for the AWS framework. A short summary of the concepts and workflow for using AWS is presented below for anyone who has not used AWS before. A more detailed guide that covers getting set up and running Loci-Stream on the AWS platform can be found on our documentation page.

General Concepts and Terms

The following bullets describe some of the relevant basic components that are commonly used on AWS.

  • Amazon Machine Image (AMI) - This is the component that represents the compiled version of the software and the operating system that it resides on.

  • Volume - This is the word used to represent the storage (hard disk, SSD) that is allocated for your machine.

  • Instance - This is a collection of an AMI, running on a specified type of machine, with a volume attached to it. This is the full “machine” that you would be starting up and shutting down in the same way you would a local computer in an office.

High-Level Workflow

If you understand how to use SSH and SCP, then the general workflow of running on AWS will be familiar to you. The high-level workflow for a user that wants to run a simulation is summarized below.

  • Create an AWS account and configure it.

  • Subscribe to the Loci-Stream AWS Marketplace page, which will grant you access to the AMIs that contain compiled versions of Loci-Stream.

  • Use an AMI to start up an instance (AMI+hardware+storage) that is sized according to your needs.

  • Once the instance is created and started up, you will then use a terminal on your local machine to SSH into the running instance.

  • Files that are necessary to start a simulation (vars and vog files) can then be tarballed up and SCP’d from your local machine to the to the instance that is running on AWS.

  • Once the files are uploaded to the instance, you can then use your SSH connection to initiate a simulation in the same way that is discussed in the Users’ guide and the tutorials.

From this point, you can monitor the progress of your simulation while SSH’d into the instance that is running Loci-Stream. Output that is written can be post-processed within the instance and then tarballed up and SCP’d down from the instance to your local machine for data analysis using Paraview or any other supported visualization tools. For a detailed description of the process of getting started with using Loci-Stream on AWS, see our AWS guide.