Automated server cluster selection for virtual machine deployment

A method, comprising:

• receiving, at a device, configuration information associated with a new virtual machine (VM) to be deployed, wherein the configuration information includes a virtual central processing unit (vCPU) parameter that indicates a number of virtual CPUs required by the new VM and a virtual memory (vMEM) parameter that indicates an amount of virtual memory required by the new VM;
• identifying, by the device, one or more network device clusters, of multiple network device clusters, assigned a value for a maximum vCPU allowed per VM that is greater than or equal to the vCPU of the new VM;
• determining, by the device, an allocation parameter that specifies an over-allocation of physical central processing units (CPUs) to virtual CPUs that is permitted within each network device cluster of the identified one or more network device clusters;
• determining, by the device, each network device cluster of the identified one or more network device clusters, whose determined allocation parameter is less than a user-defined maximum vCPU to physical CPU (pCPU) ratio, to generate a list of returned network device clusters;
• determining, for each returned network device cluster from the list of returned network device clusters, a value for available physical memory (pMEM) multiplied by a user-defined weighting parameter for each returned network device cluster;
• selecting, by the device, a network device cluster from the list of returned network device clusters having a maximum value for the pMEM multiplied by the weighted parameter;
• and
• deploying, by the device, the new VM in the selected network device cluster.

A device, comprising:

• a communication interface configured to connect to a network;
• a memory configured to store instructions;
• and
• a processing unit configured to execute the instructions stored in the memory to:
• receive configuration information associated with a new virtual machine (VM) to be deployed, wherein the configuration information includes a virtual central processing unit (vCPU) parameter that indicates a number of virtual CPUs required by the new VM and a virtual memory (vMEM) parameter that indicates an amount of virtual memory required by the new VM,
• identify one or more network device clusters, of multiple network device clusters, assigned a value for a maximum vCPU allowed per VM that is greater than or equal to the vCPU of the new VM,
• determine an allocation parameter, that specifies an over-allocation of physical central processing units (CPUs) to virtual CPUs that is permitted within each network device cluster of the identified one or more network device clusters, based on:
• 1) a maximum number of physical host failures allowed in each network device cluster,
• 2) a number of virtual CPUs currently running virtual machines in each network device cluster, and
• 3) a number of physical CPUs deployed in the multiple physical hosts of each network device cluster,
• determine each network device cluster of the identified one or more network device clusters, whose determined allocation parameter is less than a user-defined maximum vCPU to physical CPU (pCPU) ratio, to generate a list of returned network device clusters,
• determine, for each returned network device cluster from the list of returned network device clusters, a value for available physical memory (pMEM) multiplied by a user-defined weighting parameter for each returned network device cluster,
• select a network device cluster from the list of returned network device clusters having a maximum value for the pMEM multiplied by the weighted parameter, and
• cause, via the communication interface, the new VM to be deployed in the selected network device cluster.

A method, comprising:

• receiving, at a device, a virtual machine (VM) to be deployed;
• receiving, at a device, configuration information associated with the VM to be deployed, wherein the configuration information includes a virtual central processing unit (vCPU) parameter that indicates a number of virtual CPUs required by the VM to be deployed and a virtual memory (vMEM) parameter that indicates an amount of virtual memory required by the VM to be deployed;
• identifying, by the device, a plurality of network device clusters for possible VM deployment;
• applying, by the device, a weighting parameter to at least one of the plurality of network device clusters to favor selection of the at least one of the plurality of network device clusters over other of the network device clusters, wherein applying the weighting parameter comprises:
• selecting a specific weighting parameter for each of the network device clusters of the plurality of network device clusters;
• determining an amount of available physical memory for each of the network device clusters;
• and
• multiplying the specific weighting parameter with the amount of the available physical memory to generate a cluster selection value for each of the network device clusters;
• receiving, at the device, user selection of one or more network device clusters from the plurality of network device clusters to generate a disabled group of network device clusters and an enabled group of network device clusters, wherein the disabled group of network device clusters excludes at least one of the plurality of network device clusters;
• identifying, by the device, one or more network device clusters, from the enabled group of network device clusters, assigned a value for a maximum vCPU allowed per VM that is greater than or equal to the vCPU of the VM to be deployed;
• determining, by the device, an allocation parameter that specifies an over-allocation of physical central processing units (CPUs) to virtual CPUs that is permitted within each network device cluster of the identified one or more network device clusters;
• determining, by the device, each network device cluster of the identified one or more network device clusters, whose determined allocation parameter is less than a user-defined maximum vCPU to physical CPU (pCPU) ratio, to generate a list of returned network device clusters;
• selecting, by the device, a network device cluster, from the list of returned network device clusters, for deployment of the VM based on the weighting parameter applied to the at least one of the plurality of network device clusters,
• wherein selecting the network device cluster comprises:
• selecting the network device cluster of the enabled group of network device clusters having a maximum cluster selection value;
• and
• deploying the VM in the selected network device cluster.