METHOD FOR MANAGING COMPUTATIONAL RESOURCES OVER A
NETWORK
Field of the Invention
The present invention relates generally to resource management, and more particularly to method for job requests execution in peer-to-peer (P2P) for resource management architecture devoid of file sharing environment.
Background of the Invention
One of the many significant tasks of in relation to resource management involving a plurality of computing resources is to provide effectual scheduling of predefined tasks among these computing resources over a network. It is principally configured to perform job scheduling for resources on a larger scale, or as mentioned, between multiple computing resources. This particular task therefore entails additional vital tasks such as monitoring the plurality of computer resources and serves as the primary entry point for any jobs or tasks submission.
Effectively, a resource management system comprises of at least one resource management server and a plurality of resource agent components or devices. The resource agent devices are suitably configured to receive the jobs or tasks scheduled by the resource management server and subsequently execute the assigned tasks.
In order to allow carrying out the aforementioned responsibilities, the ability of file
sharing is therefore preferred within the system purely to eliminate the need of transferring files among the plurality of resources agents or devices, thereby providing time-effective task executing environment.
Peer-to-peer (P2P) technology, an approach that allows diverse connectivity between devices within a network is also profoundly known for use in effective file sharing architecture, whereby an exemplary of the system supported by peer-to-peer (P2P) technology is suitably shown in FIG 1 being the prior art. Primarily, in relation to resource management, it allows scheduling and thereby enabling the resource management server to execute tasks or jobs from another resource management server. While the existing peer-to- peer (P2P) technology provides scalable support for the file sharing ability within the resource management system, it would not operate appropriately or abruptly halted in the event that the shared file environment is discontinued or disrupted , thereby resulting to failure of the overall resource managing system. This is considered as one of the major setbacks that should be effectively prevailed when considering the file sharing system for resource managing.
Recognizing the rather prominent setback as mentioned above, the present invention has been accordingly developed to eliminate said setback and enhance the resource managing system deploying the peer-to-peer (P2P) technology.
Thus, it is the primary object of the present invention to provide a method for use in resource management of a plurality of computing resources, supported by peer-to-peer (P2P) technology within a network devoid of file sharing.
It is a further object of the present invention to provide a method for use in resource management in which there is provided a simplistic architecture including to automatically gather job execution information from respective hosts and allowing transferring of information between hosts devoid of the file sharing environment.
It is another object of the present invention to provide a method for use in resource management of a plurality of computing resources, supported by peer-to-peer (P2P) technology which is capable to replace the existing peer-to-peer (P2P) with file sharing environment in relation resource management.
These and other objects, features and advantages of the present invention will be apparent from the following detailed description of preferred embodiments thereof.
Brief Description of the Drawings
This invention will be described by way of non-limiting embodiments of the present invention, with reference to the accompanying drawings, in which:
FIG 1 provides a prior art system; and
FIG 2 provides the overall flowchart of the method in accordance with a preferred embodiment of the present invention.
Summary of Invention
The present invention discloses a method for use in managing computational resources within a network comprising the steps of:
gathering information based on job execution request from several hosts, wherein at least one host is configured for submission host, at least one host is configured for execution host and at least one host is configured for execution directory; creating a job execution directory based on the job execution request; transferring information in relation to the job execution based on said job execution directory from the submission host to the execution host; executing the job; storing the output of the executed job at the job execution directory; transferring the stored job execution directory of the executed job from the execution host to the submission host; and deleting the job execution directory and all information related to said directory.
Detailed Description of the Preferred Embodiments
In line with the above summary, the following description of a number of specific and alternative embodiments is provided to understand the inventive features of the present invention. It shall be apparent to one skilled in the art, however that this invention may be practiced without such specific details.
The present invention focuses on providing an automated stage-in/stage-out file
method for use in resource managing, which is a step ahead the prior art, such that the method effectively eliminates the need of file sharing environment in a peer-to-peer (P2P) network.
Peer-to-peer (P2P) connectivity is preferred for operating the tasks based on the method of the present invention due to the advantages such as fast retrieval of data within the interconnected devices upon prompted and aids to reduce inaccuracies in relation to outdated data.
There is provided a method supported by the peer-to-peer (P2P) network devoid of shared file system, configured based on the automated stage-in/stage out directives, whereby the system configured with the directives may include devices to perform the main tasks such as storage, scheduling and execution of jobs, respectively. Devices involve may be configured to be connected based on diverse connectivity between the devices and not dependant on a centralized server. Accordingly, any one of the devices within the network may initiate connection or communication if required.
FIG 2 provides an overall architecture of the automated file stage-in/stage-out flow configured in accordance with a preferred embodiment of the present invention.
Upon receipt of job execution information or request from remote client sites, the job execution information is gathered (SlOO) from at least three different hosts, these hosts may be in the form of servers, whereby at least one is configured for data submission host, at least one configured for execution host and at least one configured for data execution directory. It is apparent to a person skilled in the art that jobs used within the description include various
forms applications, assignments, queues and the likes. Nevertheless, it should be mentioned here that there may be provided only one server for storing all data, and sectionalized into these three categories. The main objective is to provide a centralized unit for storage of data in relation to these categories which may be retrieved or transferred from one host to another when required devoid of the file sharing environment, with the assistance of the peer-to-peer (P2P) technology. In addition, job data and information related may be updated automatically and maintained within the servers.
Ideally, the submission host primary purpose is to allow submission of jobs and information related thereto and thereby stores all job submissions whereby stored job submissions may be transferred to other hosts if prompted. Other significant roles which may be performed by this particular include scheduling jobs, reorder and reprioritize jobs. It is therefore understood that the job submission host may contain all data and input pertaining to job submissions for onward transmission or dispatched to other hosts when necessary. The second server, which is configured for execution host may contain all data and input pertaining to job executions in addition to its main function as the job execution domain for all submitted jobs. It is also noted that the data and information from this host may be transferred to another host upon prompted. The third server is configured to contain all data and input pertaining to executed jobs and applications, and therefore provided in the form of job execution directory. It is understood that all information contained within all three servers may be configured to be retrievable, updated or modified if required. The information gathered based on these servers are used as reference or basis for the file staging purposes, job execution and job monitoring.
From the above, upon prompted, the job execution directory is created (S200) for a
respective job submission. The creation of directory for the job execution request is rather essential for the method of the present invention. Each directory created is unique or specific based on the job execution request and therefore prevents mixed up of information for execution of jobs. Subsequently, upon creation of the job execution directory, all job input and information files related to the directory are transferred (S300) from the submission host to the execution host for proper execution and this step do not require file sharing environment, with the established connectivity provided by the peer-to-peer (P2P) network. From here, the selected job is suitably executed (S400) based on the input files within the job execution directory. The job output are then forwarded and thereby stored based on the created job execution directory. Upon completion of execution of the respective job, the job execution directory for that particular job together with its content are appropriately archived (S500) and transferred back (S600) from the execution host to the submission host prior to deletion (S700) of the directory and its content, as the final stage which thereby sustaining a non-sharing environment.
It is further added that the steps involved in the method in accordance with a preferred embodiment of the present invention are configured to operate automatically, or self- organized based on the job execution information gathered, upon received from client sites. It is also understood that anyone of the steps may be repeated in no particular order. With the method of the present invention which does not necessitate file sharing, the amount of file pollution or poisoning which normally exist in many peer-to-peer (P2P) file sharing systems may be effectively reduced.
From the foregoing, it would be appreciated that the present invention provides a method for providing effective and highly efficient resource management within a peer-to-
peer (P2P) network. The present invention may be modified in light of the above teachings. It is therefore understood that, within the scope of the appended claims, the invention may be practiced otherwise than as specifically described.