Someone could shed soem light on a best practice here.
I want to include the OM agent as part of a server image. But im not sure how the core id would be handled each time the server image would be deployed and have a script run to initiate the install of the agent.
My thought process would be the following:
1. Already have the agent files copied onto the server
2. initiate a script to run the install (setting the management and cert server)
3. After the install, restart the agent and perform the cert request.
My question is at what point is the core id set for the node? Will the agent check in the with managment server and then the management server give the node a core id? Or will the core id already be set thus there will be a core id mismatch?
I am not quite sure, but isn't this a situation for "opcactivate"?
As far as I know it is possible to install the agent without activating it and later running opcactivate with the name of the OM management server it should contact. Maybe you also need to run ovcert -certreq afterwards.
I am not quite sure about these fine points, but that is where I would start looking.
My 2 cents..
I think that opcactivate triggers a certificate request if the certificate configuration has not been completed.
What would be the correct command to run the install of the agent but not activate the agent? IF i can do that i would imagine i could then run another script to run the activation thus elimanting a risk of a core id mismatch each time i want to deploy the image.
and then if i were to initiate a script after deployment to run the activation it would then set the core id, correct?
Thanks for this! After looking around for this (i should have just RTFM, lol) i found out that there is a way install the OA in 'inactive mode' which does exactly what i needed. Not ever HPE support even knew about this. So thanks again.