Viewing LAMMPS Results
The run of the stock LAMMPS workflow template saves results to a persistent volume. You can access and post-process those results from a later workflow that mounts the same volume – a Jupyter notebook, for example, or an interactive shell. The example below uses the Fuzzball CLI to start a workflow that sleeps until it times out or you stop it, then connects to it to read the LAMMPS results.
$ fuzzball run --image docker://rockylinux:9 --volume <volume-name>:/data --tty -- /bin/bash
Workflow "4489d93f-b2a8-4255-bc68-b41842dbe4c5" started.
Waiting for job to start...
(fuzzball:/)$ cd /data/results/lammps_cpu
(fuzzball:/data/results/lammps_cpu)$ ls -lh
total 15M
-rw-rw-r--. 1 root 100114 15M Apr 6 18:25 6ee091f5-900b-4264-8ea4-a506326a674d-lj-my-first-input.tar.gz
(fuzzball:/data/results/lammps_cpu)$ tar -xzf 6ee091f5-900b-4264-8ea4-a506326a674d-lj-my-first-input.tar.gz
(fuzzball:/data/results/lammps_cpu)$ ls -lh lj-my-first-input/ | head -5
total 18M
-rw-rw-r--. 1 root 100114 128K Apr 6 18:25 dump.1000.jpg
-rw-rw-r--. 1 root 100114 129K Apr 6 18:25 dump.10000.jpg
-rw-rw-r--. 1 root 100114 130K Apr 6 18:25 dump.10100.jpg
-rw-rw-r--. 1 root 100114 129K Apr 6 18:25 dump.10200.jpg
(fuzzball:/data/results/lammps_cpu)$ exitReplace <volume-name> with the name of your persistent volume. To pin a specific provisioner,
use <provisioner>/<volume-name>:/data instead.
In a previous section, we showed how to create a modified workflow template that saved its LAMMPS result file to AWS S3. One method to download the results file to your workstation and view it is to use the AWS CLI.
Use the AWS CLI command aws s3 cp to download the result file to your workstation. The
command below copies the result file at S3 URI s3://<bucket>/<path...>/{{.RunName}}.tar.gz to
your working directory.
$ aws s3 cp s3://co-ciq-misc-support/results/lj-my-first-input.tar.gz .From there you can use any standard tools to view, parse, or process the output.