Project

General

Profile

BrianBockelmanReview

Development:

  • For the currently proposed solution, can you provide:
  • A component architecture diagram involving the pieces involved.
  • A diagram of the trust relationships between components.
  • How will provisioning policies be communicated from stakeholders to the system?
  • How will policy verification be done?
  • What peer efforts exist? How do you plan on collaborating with them?
  • What external product teams will you depend on? What kind of relationship do you plan on having?
  • What external services will the proposed solution rely on?
  • What monitoring is envisioned to verify provisioning policy?
  • How will resource provisioning be done?
  • What auxiliary services will be needed at remote resources? Who is in charge of these?
  • What monitoring is needed for network and data management components?
  • What are the auditing needs for this system?
  • What are the unique challenges for HEPCloud as opposed to OSG opportunistic?
  • How will job runtime environment be managed?
  • How should we verify the proposed solution works at the expected scale?
  • What considerations exist for software delivery and data management? How are the bandwidth needs calculated?
  • What are the biggest ‘missing pieces’ for the current plans and solutions.
  • What’s the short-term punchlist / TODO list? Which items are on the critical path?

Operations (If time permits):

  • What are the unique monitoring requirements needed for HEPCloud versus OSG opportunistic in FIFE?
  • How do you envision system bootstrap to proceed? I.e., how do you get from the current state to the production state?
  • What will be needed for release validation and testing?
  • How do you verify a resource is working? Excluding resources that already are in the OSG, what potential resources are not yet verified as working?
  • What sort of automated testing is seen as being needed?
  • How will HEPCloud activity be isolated / protected from other activity?
  • How will incidents / outages be handled with external resources?

Management (If time permits):

  • What are the goals and deliverables for HEPCloud?
  • What are the milestones for the next year?
  • Who are the stakeholders? What are the important deliverables for each?
  • What provisioning policies that are foreseen for the next year?
  • What are the expected resources HEPCloud should utilize in the next year?
  • What is the expected scale?
  • What are the metrics for success?