JPPF Issue Tracker
star_faded.png
Please log in to bookmark issues
bug_report_small.png
CLOSED  Bug report JPPF-485  -  Number of peer total processing threads is not propagated properly
Posted Jan 19, 2017 - updated Jan 19, 2017
icon_info.png This issue has been closed with status "Closed" and resolution "RESOLVED".
Issue details
  • Type of issue
    Bug report
  • Status
     
    		 Closed
  • Assigned to
    lolo4j
  • Progress
       
  • Type of bug
    Not triaged
  • Likelihood
    Not triaged
  • Effect
    Not triaged
  • Posted by
    lolo4j
  • Owned by
    Not owned by anyone
  • Category
    Performance
  • Resolution
    RESOLVED
  • Priority
    Normal
  • Reproducability
    Always
  • Severity
    Normal
  • Targetted for
    icon_milestones.png JPPF 5.2.5
Issue description
When a peer driver notifies its remote peer of a change in its number of nodes and/or threads the remote peer updates the properties of its peer accordingly but does not notify the associated bundler (i.e. load-balancer).

In the case of the "nodethreads" algorithm, this causes a wrong number of total processing threads for the peer driver to be computed, which in turn impairs the efficiency of load-balancing and overall performance of the grid
Steps to reproduce this issue
Setup:
  • driver1 with 1 node
  • driver1 configured "nodethreads" algorithm with multiplicator = 1
  • driver2 with 9 nodes
  • all nodes configured with 1 processingh thread
  • client app connects to driver1 only
  • client app sends 1 job with 10 tasks
==> all tasks sould be executed at once on all the available nodes, but soem nodes remain idle until availableProcessor tasks have been executed

#3
Comment posted by
lolo4j
Jan 19, 21:10
Fixed in: