August 22, 2017, Tuesday

Jump to bottom

Liya Wang edited this page Aug 22, 2017 · 5 revisions

How do we support workflow running @ TACC Stampede?
- At App level, e.g. RNAseq and Chipseq are set to run at TACC
- Data are pulled from Data Store
- Results are archived back to user's sci_data
  - Need to set up a storage system for archiving (add user name to it?)
  - URL from agave for defined outputs?
  - How do we make archiving decision automatically based on app?
  - Is it possible to optimize archiving? e.g. leave intermediate results on Stampede
  - Need to add maizecode to TACC and iPlant allocation, can we set up a cron job to clean the work folder?
    - Best to keep a separate version at de.sciapps.org
Other questions
- Can we remove the email field from the app form? If we can get email from login user
  - Change to notification button (Yes or No)
- Set a 10 MB upload limit on file: https://stackoverflow.com/questions/28605393/size-limit-for-file-upload
- For job submission, retry if Agave is not responding
  - Try three times with an interval of 3 minutes
  - Disable entire row if Agave is not responding
  - If workflow fails at any step, display error message "Step ? failed at submission, Agave might be under maintenance!"
- Output URL bugs in both diagram and history panel
  - Ignore for now, re-run workflow to fix current workflows
  - Diagram is fixed
- Add a direct link to workflows?