Nimbostratus
Employee
I assume you are referring to this document: http://www.f5.com/pdf/deployment-guides/vmware-provisioning-dg.pdf
Have you taken a look at our vSphere Plugin. http://devcentral.f5.com/Community/GroupDetails/tabid/1082223/aft/1174055/asg/2002/Default.aspx
This plugin will allow you to easily automate the F5 portion of the workflow you are describing. The plugin utilizes a rules engine one function of which is to add VMs to Pools upon VM creation. (i.e. if the VM has "Web" in the name then add to the http_pool on the LTM)
You may also want to look at this video: http://www.f5.com/videos/hybrid-cloud-application-architecture.wmv this video shows the workflow you are requesting in action. It leverages vCloud Director, and iControl API calls to accomplish the workflow you are discussing. We have a deployment guide which details how to deploy this solution here: http://www.f5.com/pdf/deployment-guides/vmware-vcloud-director-dg.pdf
One thing we have seen done to decrease the deployment time of the VM, is to have you N number of VMs Powered on, and to have +1 VM instantiated from template, but powered off. When the pool members run out of resources, you would first power on the instantiated but powered off VM, and then instantiate another VM from template. This means that you will have a pre-cloned VM ready to power on when the pool runs out of resources.
This is simply a reversal of the last two steps in the workflow you had seen in the VMware-provisioning deployment guide. instead of
Trigger event> instantiating from template, waiting for that to complete > powering that VM on,
You would
pre-create an extra VM, >trigger event> power on > instantiate from template.
Paul