Machine reassignment for Google ROADEF/EURO Challenge 2012

by - December 9, 2011
Optimization Article

Drools Planner has a new example: machine reassignment.

Problem statement

Assign each process to a machine. All processes already have an original (unoptimized) assignment. Each process requires an amount of each resource (such as CPU, RAM, …). This is more complex version of the Cloud balancing example.

The problem is defined by the Google ROADEF/EURO Challenge 2012.
Hard constraints:
  • Maximum capacity: The maximum capacity for each resource for each machine must not be exceeded.
  • Conflict: Processes of the same service must run on distinct machines.
  • Spread: Processes of the same service must be spread across locations.
  • Dependency: The processes of a service depending on another service must run in the neighborhood of a process of the other service.
  • Transient usage: Some resources are transient and count towards the maximum capacity of both the original machine as the newly assigned machine.
Soft constraints:
  • Load: The safety capacity for each resource for each machine should not be exceeded.
  • Balance: Leave room for future assignments by balancing the available resources on each machine.
  • Process move cost: A process has a move cost.
  • Service move cost: A service has a move cost.
  • Machine move cost: Moving a process from machine A to machine B has another A-B specific move cost.

Problem size
model_a1_1: 2 resources, 1 neighborhoods, 4 locations, 4 machines, 79 services, 100 processes and 1 balancePenalties with flooredPossibleSolutionSize (10^60).
model_a1_2: 4 resources, 2 neighborhoods, 4 locations, 100 machines, 980 services, 1000 processes and 0 balancePenalties with flooredPossibleSolutionSize (10^2000).
model_a1_3: 3 resources, 5 neighborhoods, 25 locations, 100 machines, 216 services, 1000 processes and 0 balancePenalties with flooredPossibleSolutionSize (10^2000).
model_a1_4: 3 resources, 50 neighborhoods, 50 locations, 50 machines, 142 services, 1000 processes and 1 balancePenalties with flooredPossibleSolutionSize (10^1698).
model_a1_5: 4 resources, 2 neighborhoods, 4 locations, 12 machines, 981 services, 1000 processes and 1 balancePenalties with flooredPossibleSolutionSize (10^1079).
model_a2_1: 3 resources, 1 neighborhoods, 1 locations, 100 machines, 1000 services, 1000 processes and 0 balancePenalties with flooredPossibleSolutionSize (10^2000).
model_a2_2: 12 resources, 5 neighborhoods, 25 locations, 100 machines, 170 services, 1000 processes and 0 balancePenalties with flooredPossibleSolutionSize (10^2000).
model_a2_3: 12 resources, 5 neighborhoods, 25 locations, 100 machines, 129 services, 1000 processes and 0 balancePenalties with flooredPossibleSolutionSize (10^2000).
model_a2_4: 12 resources, 5 neighborhoods, 25 locations, 50 machines, 180 services, 1000 processes and 1 balancePenalties with flooredPossibleSolutionSize (10^1698).
model_a2_5: 12 resources, 5 neighborhoods, 25 locations, 50 machines, 153 services, 1000 processes and 0 balancePenalties with flooredPossibleSolutionSize (10^1698).

Results
The implementation is still pretty basic, I wrote it in about 5 Sundays (including tweaking). These are the results:

The schedules found by Drools Planner (after only 5 minutes) are on average 27% better than the original schedules.

Open Science
In the name of Open Science and Open Source, all my work was public from day one (as people who follow me on Google+ or GitHub will have seen). I hope that researchers who work full-time on these sort of competitions start doing the same.

Qualification phase 1 has just ended. It has been an interesting competition so far: it has given me a good insight on the strengths and – more interestingly – the weaknesses of Drools Planner. Expect those weaknesses to be improved in future Drools Planner versions 🙂

Author

Comments are closed.