1- In DSM GUI, create a LoadBalancing root group (ex : LB).
2- Under LB group create source groups which contain computers to assign to scalability servers and one target group per scalability servers
In this example there are 2 source groups (Source_Group_1 and Source_Group_2) and 4 target groups (Scala1, Scala2, Scala3 and Scala4).
Source Group could be static groups or dynamic groups and should contain the computers which should be dispatched to the target groups.
Target Groups must be static groups.
3- Put the computers in the source groups (or evaluate the source groups if they are dynamic groups).
4- Under Control Panel/Configuration/Configuration Policy, create a new configuration policy
And put the ipaddress or name of a scalability server under
Control Panel/Configuration/Configuration Policy/Scalability server address
(Make this parameter centrally managed)
In this example configuration policy Scala1 is created and contains
Control Panel/Configuration/Configuration Policy/Scalability server address=Scala_Server1
Seal the policy and do the same for all other scalability servers
5- Apply the configuration policies to Scalability Server groups. (Do a copy/paste of a configuration policy onto a scalability server group)
The configuration policy Scala1 has been applied onto group Scala1. Every machines added in this group will receive automatically the configuration Policy Scala1 and will be linked to server Scala_Server1
6- Copy the attached files LoadBalancer.DMS and LoadBalancer.INI into C:\Program Files (x86)\CA\DSM\Bin
7- Edit file LoadBalancer.INI and put the correct values for source and targets groups :
NBSOURCE = 2
NBTARGET = 4
8- Execute the script with following command :
START /WAIT dmscript "C:\Program Files (x86)\CA\DSM\Bin\LoadBalancer.DMS"
This script generates a log "C:\Program Files (x86)\CA\DSM\logs\LoadBalancer.log"
The script will assign the computers from the source groups to target groups in an equitable manner.
After execution of the script, Groups Scala1, Scala2, Scala3 and Scala4 contains 25 computers each (there are 100 computers in the source groups)
This script could be executed regularly if there are new machines in source groups.