Skip to end of metadata
Go to start of metadata



Action Items:


Notes/ Discussion items:

  • Ron is the contact for XSEDE for IDB
  • What level of anonymization is needed for data sets received from specific SPs? John says it has to be anonymous.
  • How do we require IRB documentation? Talk to Lorna and Lizane.
  • Work with the SPs on creating anonymous data sets about their resources.
  • At the end of XSEDE we could provide a final dump of data sets. Illinois will keep these data sets as XSEDE will fade away.
  • File format: it is possible the IDB can be extended and use other formats. It seems IDB is better location than IDEALS.
  • Public vs controlled data sets: John says he does not know specifically, but he cannot imagine IDB not doing that. IDEALS separates searching for metadata and searching for documents.
  • Access to a Data Set Collection could be one of the resource allocations XSEDE could be providing through XRAS.
  • R1 vs non-R1 usage on XSEDE: it would be interesting seeing other XSEDE's data like training. We focus too much on the machines (Danna says).
  • John: What are the sorts of questions that researchers are trying to answer when they ask for data sets?. Dave: usage, interesting in doing "scheduling studies"; we can provide jobs rate data as well. Publications related to this.
  • John: how is this different from XdMod. Dave: you can get all in a junk. It is complimentary, too. Here is everything we can get credit for.
  • John: users can explore data analysis they cannot with XdMod.
  • Dave suggests offer this as a service to the different SPs.
  • IDEALS has been accepting DB files. You may need to open them with pgAdmin
  • Illinois Data Bank is about sharing data sets, too. They can handle sized of data sets. Check with them about workflow, how can they help?


  • No labels