Intelligent network storage planning within a clustered computing environment

• US 9,106,675 B2
• Filed: 06/17/2010
• Issued: 08/11/2015
• Est. Priority Date: 06/17/2010
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Alatorre, Gabriel, Butler, Eric K., Chavda, Kavita, Gopisetty, Sandeep, Murthy, Seshashayee S., Singh, Aameek
• Primary Examiner(s): Chea, Philip
• Assistant Examiner(s): Nguyen, Van Kim T
• Application Number:
  US12/817,238
• Publication Number:
  US 20110314164A1
• Time in Patent Office: 1,881 Days
• US Class Current: 1/1
• CPC Class Codes: H04L 41/0816 the condition being an adaptation, e.g. in response to network events
• CPC Class Codes: H04L 41/145 involving simulating, designing, planning or modelling of a network
• CPC Class Codes: H04L 43/0876 Network utilisation, e.g. volume of load or congestion level
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]

Abstract

Embodiments of the present invention provide an integrated host and subsystem port selection methodology that uses performance measurements combined with information about active data paths. This technique also helps in resilient fabric planning by selecting ports from redundant fabrics. In a typical embodiment, host port to storage port pairs that create a path between a host and a storage device will be identified. From these pairs, a set of host port to storage port candidates for communicate data from the host to the storage device will be identified based on a set of resiliency constraints. Then, a specific host port to storage port pair will be selected from the set based on a lowest joint workload measurement. A path will then be created between the specific host port and storage port, and data will be communicated from the host to the storage device via the path.

Claims

1. A method for intelligent network storage planning within a cloud computing environment, comprising:

defining the cloud computing environment as a network of independent cloud networks;

identifying host port to storage port pairs that create a network path between a host of one of the cloud networks and a storage device of another of the cloud networks within the cloud computing environment, each storage port in the host port to storage port pairs being used to access a single storage device;

determining a set of host port candidates having a plurality of port candidates from the host port to storage port pairs for communicating data from the host to the storage device based on a predetermined set of resiliency constraints;

calculating a joint workload measure for each of the set of host port candidates by determining, for every host port candidate in the set of host port candidates, an estimated workload for the network path between the specific host port to storage port pair, and summing a load balanced score for a the host and the storage device with the estimated workload;

selecting, for every host port candidate in the set of host port candidates, a specific host port to storage port pair based on the set of host port candidates having a lowest joint workload measure;

creating, for every host port candidate in the set of host port candidates, a path between the specific host port to storage port pair through which the data will be communicated;

iteratively repeating the identifying, determining, calculating, and selecting based on a different workload measure, the different workload measure being based on balancing a load across ports while maintaining path resiliency constraints;

dynamically attempting to identify at least one alternative path based on a result of the iteratively repeating;

creating the at least one alternative path between another host port to storage port pair;

stopping the iteratively repeating at least one of;

after a predetermined period of time or in response to a failure of the attempting.

2. A system for intelligent network storage planning within a cloud computing environment, comprising:

a bus;

a processor coupled to the bus; and

a memory medium coupled to the bus, the memory medium comprising instructions to;

define the cloud computing environment as a network of independent cloud networks;

identify, for a plurality of port candidates, host port to storage port pairs that create a network path between a host of one of the cloud networks and a storage device of another of the cloud networks within the cloud computing environment, each storage port in the host port to storage port pairs being used to access a single storage device;

select, for every port candidate of the plurality of port candidates, a specific host port to storage port pair based on the set of host port candidates having a lowest joint workload measure, wherein the joint workload measure is calculated by determining an estimated workload for the network path between the specific host port to storage port pair, and summing a load balanced score for the host and the storage device with the estimated workload;

create, for every port candidate of the plurality of port candidates, a path between the specific host port to storage port pair through which the data will be communicated;

iteratively repeat the identifying, determining, calculating, and selecting based on a different workload measure, the different workload measure being based on balancing a load across ports while maintaining path resiliency constraints;

dynamically attempt to identify at least one alternative path based on a result of the iterative repetition;

create the at least one alternative path between another host port to storage port pair; and

stop the iterative repetition at least one of;

after a predetermined period of time or in response to a failure of the attempt.

3. A computer program product for intelligent network storage planning within a cloud computing environment, the computer program product comprising a non-transitory computer readable storage medium, and program instructions stored on the computer readable storage medium, to:

define the cloud computing environment as a network of independent cloud networks;

identify host port to storage port pairs that create a network path between a host of one of the cloud networks and a storage device of another of the cloud networks within the cloud computing environment, each storage port in the host port to storage port pairs being used to access a single storage device;

determine a set of host port candidates having a plurality of port candidates from the host port to storage port pairs for communicating data from the host to the storage device based on a predetermined set of resiliency constraints;

calculate, for every host port candidate in the set of host port candidates, a joint workload measure for each of the set of host port candidates by determining an estimated workload for the network path between the specific host port to storage port pair, and summing a load balanced score for the host and the storage device with the estimated workload;

select, for every host port candidate in the set of host port candidates, a specific host port to storage port pair based on the set of host port candidates having a lowest joint workload measure;

create, for every host port candidate in the set of host port candidates, a path between the specific host port to storage port pair through which the data will be communicated;

iteratively repeat the identifying, determining, calculating, and selecting based on a different workload measure, the different workload measure being based on balancing a load across ports while maintaining path resiliency constraints;

dynamically attempt to identify at least one alternative path based on a result of the iterative repetition;

create the at least one alternative path between another host port to storage port pair; and

stop the iterative repetition at least one of;

after a predetermined period of time or in response to a failure of the attempt.

4. A method for deploying a system for intelligent network storage planning within a cloud computing environment, comprising:

providing a computer infrastructure having functionality to define the cloud computing environment as a network of independent cloud networks;

identify host port to storage port pairs that create a network path between a host of one of the cloud networks and a storage device of another of the cloud networks within the cloud computing environment, each storage port in the host port to storage port pairs being used to access a single storage device;

determine a set of host port candidates having a plurality of port candidates from the host port to storage port pairs for communicating data from the host to the storage device based on a predetermined set of resiliency constraints;

select, for every host port candidate in the set of host port candidates, a specific host port to storage port pair based on the set of host port candidates having a lowest joint workload measure, wherein the joint workload measure is calculated by determining an estimated workload for the network path between the specific host port to storage port pair, and summing a load balanced score for the host and the storage device with the estimate workload;

create, for every host port candidate in the set of host port candidates, a path between the specific host port to storage port pair through which the data will be communicated;

iteratively repeat the identifying, determining, calculating, and selecting based on a different workload measure, the different workload measure being based on balancing a load across ports while maintaining path resiliency constraints;

dynamically attempt to identify at least one alternative path based on a result of the iterative repetition;

create the at least one alternative path between another host port to storage port pair; and

stop the iterative repetition at least one of;

after a predetermined period of time or in response to a failure of the attempt.