Hi guys,
I need to do automatic VM deploy when a threshold exceed.
Ex. If VM's had CPU on 90% per 20 min create another VM. ( I have a LVS to adjust connections)
Someone have docs or links to help me?
tks
Hi!
Great Use-Case: 'The self-healing datacenter'
I don't know ready-to-use examples, but some ideas:
Deploying a new VM (e.g. based from a template) within a workflow is no challenge, and there are a lot of examples out there...
http://www.vcoteam.info/development/how-to-deploy-vms-form-template-in-orchestrator.html
http://communities.vmware.com/message/1733734#1733734
http://communities.vmware.com/message/1835915
(and all the Links in there)
The more challenging part is, how to trigger that workflow based on CPU utilization?
1. Do the monitoring manually within another workflow. For that you can access the vCenter Performance Data in the script, see here:
http://communities.vmware.com/message/1646292
http://communities.vmware.com/message/2029115
For more background information: http://vmware-forum.de/viewtopic.php?p=123416#123416 (It's in German, but the referenced links are in English
2. Trigger the workflow based on vCenter Alarms (so that vCenter does the monitoring). You can use the SNMP-Plugin for that.
http://blogs.vmware.com/orchestrator/2011/09/snmp-plug-in-integration-with-vcenter.html
3. Use an external Monitoring System, and trigger the Workflow via AMQP, SNMP, direct API calls...
An inspiring article how to do this with vCenter Operations: http://www.vcoportal.de/2012/05/integrate-vcops-and-vco/
(See the links at the end of the article!)
For an smooth start I recommend the 2. way. Configure vCenter alarms properly (than it's not your job to deal with the performance counter), and use the SNMP-Plugin to kick-off the Deploy Workflow.
Cheers,
Joerg
Tks, This tips will help me