Dynamic Hierarchical Reserved Resource Allocation

Abstract

A method includes receiving a request to allocate an instantiation of a network function and information indicative of resource needs of the instantiation. The resource needs include at least one resiliency requirement. The method includes computing a resource map comprising a global tier and a regional tier and comparing the resource needs with the resource map to determine an allocation solution. The method also includes, based on the allocation solution, allocating resources to the instantiation. The resources include a first resource of the global tier and a second resource of the regional tier.

What is claimed: 
     
         1 . A cloud orchestrator comprising:
 a network connection for connecting to a cloud network;   a processor communicatively coupled to the network connection; and   memory storing instructions that cause the processor to effectuate operations, the operations comprising:   computing a resiliency score for each of a plurality of allocation solutions for an instantiation of a network function in the cloud network, wherein the resiliency score is based on a local resiliency and a geographic resiliency of resources in the plurality of allocation solutions;   based on the resiliency score and a resiliency requirement of the instantiation of the network function, allocating resources to the instantiation of the network function, wherein the resources comprise a first resource of a global tier and a second resource of a regional tier;   responsive to the allocating the resources to the instantiation, periodically:
 updating a resource map of the cloud network to generate a predictive heat map of free resources of the cloud network; and 
 determining a plurality of potential reallocation solutions for mitigating potential resource failures affecting the network function according to the predictive heat map of free resources of the cloud network; 
   detecting a first resource failure affecting the network function; and   implementing a first reallocation solution of the plurality of potential reallocation solutions responsive to the detecting the first resource failure affecting the network function.   
     
     
         2 . The cloud orchestrator of  claim 1 , wherein the operations further comprise comparing resource requirements with the resource map to determine the plurality of allocation solutions. 
     
     
         3 . The cloud orchestrator of  claim 2 , wherein the resource requirements comprise a geographic preference, and wherein the second resource is selected based on the geographic preference. 
     
     
         4 . The cloud orchestrator of  claim 1 , wherein the operations further comprise computing the resource map, wherein the resource map includes the first resource of the global tier and the second resource of the regional tier. 
     
     
         5 . The cloud orchestrator of  claim 1 , wherein the operations further comprise receiving a request to allocate the instantiation of the network function, wherein the request comprises information indicative of resource requirements of the instantiation. 
     
     
         6 . The cloud orchestrator of  claim 5 , wherein the resource requirements include the resiliency requirement of the instantiation, and wherein the resiliency requirement includes a datacenter. 
     
     
         7 . The cloud orchestrator of  claim 1 , the operations further comprising:
 determining that a new datacenter has been created within the regional tier; and   controlling relocation of the instantiation from the second resource to a third resource of the new datacenter.   
     
     
         8 . The cloud orchestrator of  claim 1 , wherein the resource map further comprises an edge tier, and wherein a first latency of the edge tier is lower than a second latency of the regional tier. 
     
     
         9 . A method, comprising:
 computing, by a processing system including a processor, a resiliency score for each of a plurality of allocation solutions for an instantiation of a network function in a cloud network, wherein the resiliency score is based on a local resiliency and a geographic resiliency of resources in the plurality of allocation solutions;   based on the resiliency score and a resiliency requirement of the instantiation of the network function, allocating, by the processing system, resources to the instantiation of the network function;   responsive to the allocating the resources to the instantiation, periodically:
 updating, by the processing system, a resource map of the cloud network to generate a predictive heat map of free resources of the cloud network; and 
 determining, by the processing system, a plurality of potential reallocation solutions for mitigating potential resource failures affecting the network function according to the predictive heat map of free resources of the cloud network; 
   detecting, by the processing system, a first resource failure affecting the network function; and   implementing, by the processing system, a first reallocation solution of the plurality of potential reallocation solutions responsive to the detecting the first resource failure affecting the network function.   
     
     
         10 . The method of  claim 9 , further comprising comparing, by the processing system, resource requirements with the resource map to determine the plurality of allocation solutions. 
     
     
         11 . The method of  claim 10 , wherein the resource requirements further comprise an affinity rule, non-affinity rule, or a combination thereof. 
     
     
         12 . The method of  claim 10 , further comprising receiving, by the processing system, a request to allocate the instantiation of the network function, wherein the request comprises information indicative of the resource requirements. 
     
     
         13 . The method of  claim 10 , wherein the resource requirements include the resiliency requirement of the instantiation, and wherein the resiliency requirement includes a datacenter. 
     
     
         14 . The method of  claim 10 , further comprising:
 comparing, by the processing system, the resource requirements to a second resource need of a second network function; and   identifying, by the processing system, a first allocation solution and a second allocation solution based on the resource requirements, the second resource need, and the resource map.   
     
     
         15 . The method of  claim 14 , wherein the first allocation solution has a first reliability score that is higher than a second reliability score of the second allocation solution. 
     
     
         16 . The method of  claim 14 , wherein the first allocation solution satisfies a geographic preference associated with the instantiation. 
     
     
         17 . The method of  claim 9 , wherein the resources comprise a first resource of a global tier and a second resource of a regional tier. 
     
     
         18 . The method of  claim 17 , wherein the global tier comprises a global datacenter and the regional tier comprises a regional datacenter. 
     
     
         19 . A non-transitory computer readable storage medium storinginstructions that cause a processor executing the instructions to effectuate operations, the operations comprising:
 computing a resiliency score for each of a plurality of allocation solutions for an instantiation of a network function in a cloud network;   based on the resiliency score and a resiliency requirement of the instantiation of the network function, allocating resources to the instantiation of the network function, wherein the resources comprise a first resource of a global tier and a second resource of a regional tier;   responsive to the allocating the resources to the instantiation, periodically:
 updating a resource map of the cloud network to generate a predictive heat map of free resources of the cloud network; and 
 determining a plurality of potential reallocation solutions for mitigating potential resource failures affecting the network function according to the predictive heat map of free resources of the cloud network; 
   detecting a first resource failure affecting the network function; and   implementing a first reallocation solution of the plurality of potential reallocation solutions responsive to the detecting the first resource failure affecting the network function.   
     
     
         20 . The non-transitory computer readable storage medium of  claim 19 , wherein the resiliency score is based on a local resiliency and a geographic resiliency of resources in the plurality of allocation solutions.