This is a suite of simple tools to help the user to Handle Darshan Data without being familiar exactly with specific commands.
- Connect to the system with
ssh -Y
-
Edit the script open_darshan.sh and declare for the variable darshan_path the path where the logs are saved
-
To load Darshan data for your last experiment on the same day, execute
./open_darshan.sh
- To load darshan data from specific job, execute
./open_darshan.sh job_id
The PDF file will be created in the folder
experiments/year/month/executable/ (v0.2)
or
experiments/executable/year/month/ (development)
-
Available with release v0.3 and afterwards
-
Dependency on PDFjam software
-
Use the executable in the bin folder. We do not support this software
-
Edit the script compare_darshan.sh and declare for the variable darshan_path the path where the logs are saved
-
To compare the Darshan data of two jobs with job_id1 and job_id2, execute
./compare_darshan.sh job_id1 job_id2
Example of one single PDF:
{% for relea in site.github.releases %}
Version: {{ relea.tag_name }}
Date: {{ relea.published_at }}
Changelog:
{{ relea.body }}
[Download]({{ relea.tarball_url }})
{% endfor %}
The scripts were tested with CRAY-XC40 ShaheenII and Darshan v2.3.1
- Automatic organization into folders (v0.2 and afterwards)
- Adapt with newer Darshan version