Allocation Time and Congestion Handling

From ReddNet
Revision as of 09:01, 21 April 2008 by Sheldonp (talk | contribs) (New page: * problem is that there is a big overhead to checking all the allocation on a depot or a set of depots. * so want to minimize how often this is done on a recurring basis * so how short a t...)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search
  • problem is that there is a big overhead to checking all the allocation on a depot or a set of depots.
  • so want to minimize how often this is done on a recurring basis
  • so how short a time is "just short enough"?
  • at depot level it will be best effort, real congestion handling should be done at higher level - at LoCI tools or L-Store
  • for the next 6 months, our plan:
    • see what "short enough" is, try a month at beginning, maybe adjust as needed
    • work at LoCI/L-Store levels to do this right
    • implement a manager for something like a really short term duration use case as an interim solution (for example for the Condor checkpoint file use case).