Using the 'Bare Metal' target functionality to add physical machines for Cloud migration plans.

Document created by dominic.grigg Expert on Mar 7, 2019
Version 1Show Document
  • View in full screen mode
  • Purpose  of the Bare Metal target:

 

The Bare Metal Hypervisor target exists solely for the purpose of running Physical Machine to Cloud migration plans. This is done by creating a 'fake' VM in the On-prem env to represent the physical machine. This is done by using initial metrics supplied in a json file that is referenced when adding the target. Later WMI or SNMP targets are created to get actual utilisation metrics from the physical machine. This functionality has existed since 6.2.3. None of the usual on-prem functionality for optimisation can be used with these targets as they are of course physical machines and so cannot be migrated or scaled in as real VMs are. 

 

  • Creating a Bare Metal target:  

       

To create a physical machine target with capacity information we require a JSON file to be built for each group of physical servers. The JSON file needs to be in the format of the following example:  

        

{"hosts": [

{"entityId" : "id2","displayName": "Bob","osName" : "Windows", "numOfCPU" : "4", "cpuSpeedMhz" : "2600", "memsizeGB" : "8", "diskSizeGB" : "80", "ipAddresses": ["10.16.172.54"]},

{"entityId" : "id3","displayName": "Fred","osName" : "Linux", "numOfCPU" : "2", "cpuSpeedMhz" : "2800", "memsizeGB" : "2", "diskSizeGB" : "60", "ipAddresses": ["10.16.172.55"]},

{"entityId" : "id4","displayName": "Sid","osName" : "Linux", "numOfCPU" : "2", "cpuSpeedMhz" : "2400", "memsizeGB" : "4", "diskSizeGB" : "50", "ipAddresses": ["10.16.172.56"]},

{"entityId" : "id5","displayName": "Bill","osName" : "Windows", "numOfCPU" : "4", "cpuSpeedMhz" : "2800", "memsizeGB" : "4", "diskSizeGB" : "100", "ipAddresses": ["10.16.172.57"]}

]}

       

For testing JSON files can be generated using an on-line JSON file editor tool such as JSON Editor Online - view, edit and format JSON online 

The created JSON file needs to accessible either locally to the Turbonomic instance (ie. scp'ed to /tmp on the instance or on a network share accessible by the Turbonomic instance). 

Enter a target name for the instance and then the location of the JSON file as per the screenshot below: 

 

 

Note three slashes used for the filename. https uses only two slashes as usual. 

 

Once the target is validated Turbonomic will create pseudo-VMs within the market using the JSON file and group associated with those 'VM's. to get utilisation data needed for a migration plan it's necessary to create WMI/SNMP Guest OS Process targets as per the instructions in the Target configuration guide. 

 

Once a pseudo-VM has been created from the JSON file it should be possible to find the VM in the inventory. Capacity and usage will look something like this: 

 



 

Notice that many of the capacity values are set to PB or Phz. This is by design and are set very high to accommodate any utlisation of the pseudo-VM. As explained by Engineering these PB and Phz capacities are for consumed commodities from the fake, non-existent host that this pseudo-VM 'resides' on.  

Attachments

    Outcomes