Information requested on how load is distributed across a molecule

Document created by rich_patterson Employee on Jul 16, 2014Last modified by mike_aronson on Sep 23, 2016
Version 2Show Document
  • View in full screen mode
Assuming your molecule is in good health, the head node will be responsible for administration, scheduling, and distributing processes across the nodes in your molecule.

Because the scheduling and load balancing itself can consume hardware resources, processes are typically sent to the child nodes first, before it executes processes itself.

The load distribution performed by the head node, does not take into account other, non-AtomSphere processing on a particular server.  For example, the molecule does not monitor CPU load of a particular server when distributing processes.  Therefore, if there are non-AtomSphere applications consuming resources on individual processors, this may impact your CPU utilization.

3 people found this helpful