Selection of a resource in a distributed computer system

Abstract

A method for selecting a resource from a plurality of resources includes determining a score for that resource on the basis of a stochastic property of that resource. An interval corrsesponding to that resource is then defined to have an extent that depends on the score. A particular resource is then selected by generating a random number and selecting that resource when the random number falls within the interval.

1 .- 26 . (canceled)  
   
   
       27 . A method for selecting a receiving adaptor in connection with communication between storage devices, comprising: 
 assigning a score to each of a plurality of possible receiving adaptors;    defining an interval for each of said possible receiving adaptors, wherein each interval varies according to a corresponding score of each of said plurality of possible receiving adaptors;    generating a random number; and    selecting a particular receiving adaptor from said plurality of possible receiving adaptors, wherein said particular receiving adaptor has an interval corresponding to said random number.    
   
   
       28 . A method, according to  claim 27 , further comprising: 
 establishing communication with said particular receiving adaptor.    
   
   
       29 . A method, according to  claim 27 , wherein generating a random number includes generating a uniformly distributed random number.  
   
   
       30 . A method, according to  claim 27 , wherein each of said scores varies according to an inverse workload value of each of said possible receiving adaptors.  
   
   
       31 . A method, according to  claim 30 , further comprising: 
 summing each of the inverse workload values to obtain a normalizing factor.    
   
   
       32 . A method, according to  claim 31 , wherein each of said intervals for each of said possible receiving adaptors is a portion of said normalizing factor having a size that varies according to a corresponding one of each of said inverse workload values.  
   
   
       33 . A method, according to  claim 32 , wherein each of said inverse workload values is obtained from a workload table.  
   
   
       34 . A method, according to  claim 33 , further comprising: 
 locally caching the workload table.    
   
   
       35 . A method, according to  claim 33 , wherein said workload table is updated each time data is sent from one of the storage devices to another one of the storage devices.  
   
   
       36 . Computer software, provided in a computer-readable medium, for selecting a receiving adaptor in connection with communication between storage devices, comprising: 
 executable code that assigns a score to each of a plurality of possible receiving adaptors;    executable code that defines an interval for each of said possible receiving adaptors, wherein each interval varies according to a corresponding score of each of said plurality of possible receiving adaptors;    executable code that generates a random number; and    executable code that selects a particular receiving adaptor from said plurality of possible receiving adaptors, wherein said particular receiving adaptor has an interval corresponding to said random number.    
   
   
       37 . Computer software, according to  claim 36 , further comprising: 
 executable code that establishes communication with said particular receiving adaptor.    
   
   
       38 . Computer software, according to  claim 36 , wherein executable code that generates a random number generates a uniformly distributed random number.  
   
   
       39 . Computer software, according to  claim 36 , wherein each of said scores varies according to an inverse workload value of each of said possible receiving adaptors.  
   
   
       40 . Computer software, according to  claim 39 , further comprising: 
 executable code that sums each of the inverse workload values to obtain a normalizing factor.    
   
   
       41 . Computer software, according to  claim 40 , wherein each of said intervals for each of said possible receiving adaptors is a portion of said normalizing factor having a size that varies according to a corresponding one of each of said inverse workload values.  
   
   
       42 . Computer software, according to  claim 41 , wherein each of said inverse workload values is obtained from a workload table.  
   
   
       43 . Computer software, according to  claim 42 , further comprising: 
 executable code that locally caches the workload table.    
   
   
       44 . Computer software, according to  claim 42 , further comprising: 
 executable code that updates said workload table is updated each time data is sent from one of the storage devices to another one of the storage devices.    
   
   
       45 . A data storage system, comprising: 
 a common memory;    a host adaptor, coupled with the common memory, that has in interface for a host;    a mass storage device coupled with the common memory;    a remote adaptor coupled with the common memory, wherein the remote adaptor causes data written to the mass storage device to be transmitted to an other storage device having a plurality of possible receiving adaptors that receive the data, and    computer software, provided in a computer-readable medium, that selects a particular one of the possible receiving adaptors, said computer software including executable code that assigns a score to each of the plurality of possible receiving adaptors, executable code that defines an interval for each of said possible receiving adaptors, wherein each interval varies according to a corresponding score of each of said plurality of possible receiving adaptors, executable code that generates a random number, and executable code that selects the particular receiving adaptor from said plurality of possible receiving adaptors, wherein said particular receiving adaptor has an interval corresponding to said random number.    
   
   
       46 . A data storage system, according to  claim 46 , wherein each of said scores varies according to an inverse workload value of each of said possible receiving adaptors.