ON A WORKLOAD TUNED SERVER

BACKGROUND

[0001] Computing resources, such as workload tuned servers, are accessed over a network by a user. By accessing the computing resources over the network, the user can access the computing resources remotely to execute computing tasks. Further, the computing resources may include an infrastructure such as a data center that is designed to support remote computing services. The data center includes a number of workload tuned servers to execute computing tasks sent by the user over the network.

BRIEF DESCRIPTION OF THE DRAWINGS

[0002] The accompanying drawings illustrate various examples of the principles described herein and are a part of the specification. The examples do not limit the scope of the claims.

[0003] Fig. 1 is a diagram of an example of a system for deploying computing tasks on workload tunes servers, according to one example of principles described herein.

[0004] Fig. 2A is a flowchart of a method for optimizing a computing task for deployment on a workload tuned server, according to one example of principles described herein.

[0005] Fig. 2B is a flowchart of a method for optimizing a computing task for deployment on a workload tuned server, according to one example of principles described herein. [0006] Fig. 3 is a diagram of an example for provisioning and deploying computing tasks on workload tuned servers, according to one example of principles described herein.

[0007] Fig. 4 is a diagram of an example of a computing task deployment service, according to one example of principles described herein.

[0008] Fig. 5 is a diagram of an example of a provisioning request, according to one example of principles described herein.

[0009] Fig. 6 is a diagram of an example of an infrastructure controller, according to one example of principles described herein.

[0010] Fig. 7 is a diagram of an example of a data center, according to one example of principles described herein.

[001 1 ] Fig. 8 is a diagram of an example of an optimizing system, according to one example of principles described herein.

[0012] Throughout the drawings, identical reference numbers designate similar, but not necessarily identical, elements.

DETAILED DESCRIPTION

[0013] To execute computing tasks sent by the user over the network, a data center uses workload tuned servers to execute the computing tasks. A workload tuned server is designed specifically to provide the greatest throughput based on a workload type. For example, a workload tuned server may be designed to execute a computing task such as a high performance computing task. As a result, the workload tuned server utilizes computing resources more efficiently and effectively when executing the high performance computing task.

[0014] Further, a workload tuned server may be designed to execute a specific workload type. However, if the workload tuned server executes a workload type that is not specific to the workload tuned server, the throughput of the workload tuned server is compromised. As a result, the computing task may be delayed in execution since the workload tuned server is not designed to optimally execute the computing task. [0015] The principles described herein inciude a system and a method for optimizing a computing task for deployment on a workload tuned server. In one example the method includes assigning at least one workload type to a computing task, matching the at least one workload type to at least one workload tuned sever, sending a provisioning request with the at least one workload type to the at least one workload tuned server, and deploying the computing task on the at least one workload tuned server to optimize the computing task. Such a method allows workload tuned servers to provide the greatest throughput for the workload type the work tuned servers are executing. As a result, the workload tuned servers utilize computing resources more efficiently and effectively for lower powe usage and better performance.

[0016] In the present specification and in the appended claims, the term "workload tuned server" is meant to be understood broadly as a computing device designed to process categories of computing tasks. Further a workload tuned server may be designed for processing performance with a general or specific type of computing task. Still further, the workload tuned server may be designed to optimize throughput on specific parts of a computing task.

[0017] in the present specification and in the appended claims, the term "deploy" or combinations thereof are meant to be understood broadly as the process of installing, executing, or processing a computing task. In one example, a computing task may be deployed on a workload tuned server.

[0018] In the present specification and in the appended claims, the term "computing task" is meant to be understood broadly as computer readable code, assembly language code, or other machine readable instructions which may be executed by a processor. In one example, a computing task may be an application, a part of the application, a script, or combinations thereof.

[0019] Further, as used in the present specification and in the appended claims, the term "a number of" or similar language is meant to be understood broadly as any positive number comprising 1 to infinity; zero not being a number, but the absence of a number.

[0020] In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present systems and methods. St will be apparent, however, to one skilled in the art that the present apparatus, systems, and methods may be practiced without these specific details. Reference in the specification to "an example" or similar language means that a particular feature, structure, or characteristic described in connection with that example is included as described, but may not be included in other examples.

[0021] Referring now to the figures, Fig.1 is a diagram of an example of a system (100) for deploying computing tasks on workload tunes servers, according to one example of principles described herein. As will be described below the system (100} includes a computing task deployment service (101 ), an infrastructure controller (102), and a workload tuned server (104). Further, the computing task deployment service (101 ) is presented to a user. This allows a user to select a computing task to be deployed on a workload tuned server.

[0022] In one example of Fig. 1 , a number of computing devices may be used for deploying computing tasks (1 10) on workload tuned servers (104). in this example, a computing task deployment service (101 ) is provided to a user. In one example, the computing task deployment service (101 } provides a user with access to computing resources, such as computing tasks (1 10). In one example, the user may select a computing task (1 10) via a user interface module (106). In this example, the computing tasks (1 10) are associated with workload types (1 12). Further, the workload types (1 12) indicate which type of workload tuned servers (104) may be used to deploy the computing tasks efficiently and effectively. As will be described in other parts of this specification by selected a computing task (1 10), the computing tasks (1 10) are deployed on workload tuned servers (104).

[0023] Further, the computing task deployment service (101 ) may be used for a number of different services such as software as a service (SaaS), platform as a service (PaaS), infrastructure as a service (iaaS), hardware as a service (HaaS), automated product deployment services, digital storage services, security computing resources services, computer rendering services, deploying various type of computing task, other services, or combination thereof. [0024] Further, the computing task deployment service (101 ) includes a provisioning request module (108). In one example, the provisioning request module (108) creates and sends a provisioning request to the infrastructure controller (102). More information about the provisioning request will be described in later parts of this specification.

[0025] As mentioned above, the optimizing system (100) includes an infrastructure controller (102). In one example, the computing task deployment service (101 ) is connected to an infrastructure controller (102). As a result, information may be exchanged between the infrastructure controller (102) and the computing task deployment service (101 ). Sn one example, the

infrastructure controller (102) provisions workload tuned servers (104) to be used by a service.

[0026] In one example, the infrastructure controller (102) includes a detection module (1 14), a provisioning module (1 15), and tuned server database (1 16). In this example, the workload tuned server database (1 18) includes server entry A (1 18) and server entry B (122) with assigned workload type A (120) and an assigned workload type B (124) respectively. In this example, the provisioning module (1 15) receives provisioning requests from the computing task deployment service (101 ) that include a number of workload types (1 12). Further, the detection module (1 14) detects workload tuned servers (1 14) that match the workload types (1 12) in the provisioning request by referencing the server entries (1 18 and 1 12) in the workload tuned server database (1 16).

[0027] As mentioned above, the optimizing system (100) includes workload tuned servers (104). Sn one example, the workload turned servers (104) are tuned to perform efficiently and effectively with specific workload types. Further, the workload tuned servers (104) are provisioned by the infrastructure controller (102) to execute computing tasks on the computing task deployment service (101 ). More information about provisioning will be described in more detail in later parts of this specification.

[0028] Fig. 2A is a flowchart of a method for optimizing a computing task for deployment on a workload tuned server, according to one example of principles described herein. In one example, the method (200) includes assigning (201 ) at least one workload type to a computing task, matching (202) the at least one workload type to at least one workload tuned sever, sending (203) a provisioning request with the at least one workload type to the at least one workload tuned server, and deploying (204) the computing task on the at least one workload tuned server to optimize the computing task.

[0029] As mentioned above, the method (200) includes assigning

(201 ) at least one workload type to a computing task. In one example, computing tasks have various characteristics that make some computer architectures more suitable to compute the computing tasks more effectively. For example, some computing tasks, such as computer image rendering, may include several machine readable instructions that are processed in parallel. In this example, a computer architecture that includes multiple processors may provide better throughput over an architecture with a single processor. In this example, the computing task may have a workload type such as a parallel processing workload type. As a result, the computing task may is assigned at least one workload type. In other examples, several workload types may be assigned to the computing task.

[0030] As mentioned above, the method (200) includes matching

(202) the at least one workload type to at least one workload tuned sever. In one example, a workload tuned server may be able to execute some workload types more efficiently than other workload types. Sn this example, if a workload type is matched properly to a workload tuned server, the computing task executing on the workload tuned server experiences an enhanced performance. Further, once a workload tuned server is matched to a computing task, the workload tuned server can be provisioned to execute the computing task.

[0031] Further, in one example, one workload type is matched to one workload tuned server. In another example, several workload types are match to one workload tuned sever. In yet another example, one workload type is matched to several workload tuned servers.

[0032] As mentioned above, the method (200) includes sending (203) a provisioning request with the at least one workload type to the at least one workload tuned server, In one example, a provisioning request includes a workload type. Further, the provisioning request may include specifications for computing infrastructure such as number of processors, random access memory (RAM) size, memory size, or combinations thereof. Further, including a workload type in a provisioning request improves the likelihood of a computing task being matched to the computing task's optimal computer architecture of a workload tuned server. As mentioned above, once a workload tuned server is matched to the computing task, the workload tuned server can be provisioned to execute the computing task.

[0033] As mentioned above, the method (200) includes deploying (204) the computing task on the at least one workload tuned server to optimize the computing task. As mentioned above, a workload tuned server is tuned for the workload type of the computing task. In one example, the computing task is deployed on a workload tuned server that is matched to the computing task's workload type. Further, by deploying the computing tasks on the workload tuned server, the computing task is executed by the workload tuned server. As a result, deploying (204) the computing task on the workload tuned server that is matched to the workload type allows the computing task to be optimized.

[0034] Fig. 2B is a flowchart of a method for optimizing a computing task for deployment on a workload tuned server, according to one example of principles described herein. In one example, the method (250) includes receiving (251 ) a request to deploy a computing task on at least one workload tuned server, assigning (252) at least one workload type to a computing task, storing (253) the at least one workload type in a database, matching (254) the at least one workload type to at least one workload tuned sever, sending (255) a provisioning request with the at least one workload type to the at least one workload tuned server, provisioning (258) the at least one workload tuned server based on the provisioning request, integrating (257) at least one service with the at least one workload tuned server, and deploying (258) the computing task on the at least one workload tuned server to optimize the computing task.

[0035] As mentioned above, the method (250) includes receiving (251 ) a request to deploy a computing task on at least one workload tuned server. In one example, a user creates a request to deploy a computing task on at least one workload tuned server. As a result, the request to deploy a computing task on at least one workload tuned server is received. In another example, several users may create a request to deploy a computing task on at least one workload tuned server. As a result, the request to deploy a computing task on at least one workload tuned server is received.

[0038] As mentioned above, the method (250) includes storing (253) the at least one workload type in a database. In one example, a computing task is assigned a workload type. In this example, the workload type assigned to the computing task is stored in a database. In one example, the database may be the workload tuned server database of Fig. 1 . In another example, the database may be another database of than the workload tuned server database of Fig. 1 .

[0037] As mentioned above, the method (250) includes provisioning (256) the at least one workload tuned server based on the provisioning request. In one example, provisioning at least one workload tuned server based on the provisioning request includes receiving the provisioning request with the at least one workioad type, in one example, the provisioning request may be received by the computing task deployment service's provisioning request module of Fig. 1. In another example, the provisioning request may be received by the infrastructure controllers provisioning module of Fig. 1 . In yet another example, the provisioning request may be received by any appropriate mechanism.

[0038] In keeping with the given example, provisioning at least one workload tuned server based on the provisioning request includes detecting the at least one workload tuned server that matches the at least one workload type. In one example, a workioad tuned server may be able to execute some workload types more efficiently than other workload types. In this example, if a workload type is detected to match properly to a workload tuned server, the computing task executing on the workload tuned server experiences an enhanced performance. Once a workioad tuned serve is matched to a computing task, the workload tuned server can be provisioned to execute the computing task. [0039] As mentioned above, the method (250) includes integrating (257) at least one service with the at least one workload tuned server, in one example, integrating at least one service with the at least one workload tuned server includes receiving an integrating request for the at least one service, in one example, services may include a SaaS, a PaaS, a iaaS, a HaaS, automated product deployment services, digital storage services, security computing resources services, computer rendering services, deploying various type of computing task, other services, or combination thereof.

[0040] Further, integrating at least one service with the at least one workload tuned server includes sending the integrating request to the workload tuned server. Further, integrating at least one service with the at least one workload tuned server includes deploying the at least one service on the workload tuned server.

[0041] Fig. 3 is a diagram of an example for provisioning and deploying computing tasks on workload tuned servers, according to one example of principles described herein. As will be described in Fig. 3, two users request to deploy a computing task, in the example of Fig. 3, the computing tasks are deployed on a workload tuned server.

[0042] in the example of Fig. 3, users (301 ) subscribe to a computing task deployment service (302). For example, user A and user B subscribe for a computing task deployment service (302). In this example, user A makes a request for a computing task, such as a online collaboration tool web computing task, to be deployed, in keeping with the given example, User B makes a request for a computing task, such as weather forecasting computing task, to be deployed. As a result, the users (301 ) request to deploy a computing task (308).

[0043] In this example, the computing task deployment service (302) receives the users' requests to deploy a computing task (306). The computing task deployment service sends a provisioning request (308) to an infrastructure controller (303). In this example, the computing task deployment service sends a provisioning request for user A and a provisioning request for user B. Further, the provisioning request for user A includes a workload type. Further, the provisioning request for user B includes another workload type. More information about the computing task deployment service (302) will be described in more detail later on in this specification.

[0044] Further, the infrastructure controller (303) detects (308) workload tuned serves that match the workload type. In one example, the workload tuned servers may be bare metal workload tuned servers. In this example, after the infrastructure controller (303) detects (308) workload tuned servers that match the workload type, this information is received by a data center (304). In this example, the data center (309) returns (309) details of the workload tuned servers that match the workload type to the infrastructure controller (303).

[0045] In keeping with the given example, the infrastructure controller (303) sets up (310) workload tuned server details in a boot server (305). In one example, this information is received by the boot server (305). Further, the infrastructure controller (303) defects (31 1 ) a workload tuned server detected by the boot sever (305). In one example, this information is received by the data center (304).

[0046] Further, a request (312) for a file image sent from the data center (304) is received by the boot server (305). In this example, the boot server (305) transfers (313) the image file to the data center (304). Next, the infrastructure controller (303) returns (314) the provisioned workload tuned server details (314) to the computing task deployment service (302). Finally, the computing task deployment service (302) deploys (315) the computing task on the workload tuned server. In this example, the workload tuned sever may be located in the data center (304).

[0047] Fig. 4 is a diagram of an example of a computing task deployment service, according to one example of principles described herein. As mentioned above, a computing task deployment service (400) is provided to a user. In one example, the computing task deployment service (400) provides a user with access to computing tasks (406). Further, the computing tasks (406) may be selected, by the user, to be deployed on a workload tuned server. [0048] In the example of Fig, 4, the user may select a computing task (408) via a user interface module (402), In this example, the computing tasks (406) include a weather forecasting computing task (408) and an online collaboration tool web computing task (410). Further, the computing tasks (406) are associated with workload types (412). In this example, the weather forecasting computing task (408) is associated with a high performance computing workload type (414). In keeping with the given example, the online collaboration tool web computing task (410) is associated with a web computing task workload type. Further, the workload types (412) indicate which type of workload tuned servers may be used to deploy the computing tasks efficiently and effectively. By selecting computing tasks (406), the computing tasks (406) are deployed on workload tuned servers.

[0049] Further, the computing task deployment service (400) includes a provisioning request module (404). As mentioned above, the provisioning request module (404) creates and sends a provisioning request to an infrastructure controller.

[0050] Fig. 5 is a diagram of an example of a provisioning request, according to one example of principles described herein. As mentioned above, a provisioning request may include a number of specifications for workload tuned servers. Turning to Fig. 5, a provisioning request (500) may specify a number of processors (502) that are to be used for the provisioning request (500). in one example, the number of processors (502) may specify a minimum number of processors to be used for the provisioning request (500). For example, the numbe of processors (502) for the provisioning request may specify six processors are to be used for the provisioning request (500). In another example, the number of processors (502) for the provisioning request may specify two processors are to be used for the provisioning request (500).

[0051] In keeping with the given example, the provisioning request (500) may specify a RAM size (504) that is to be used for the provisioning request (500). In one example, the RAM size (504) may be fifty Kilobytes (KB). In another example, he RAM si2e (504) may be ten Gigabytes (GB). in yet another example, the RAM size (504) may be any appropriate size for the provisioning request (500).

[0052] In keeping with the given example, the provisioning request (500) may specify a workload type (508) that is to be used for the provisioning request (500). in one example, the workload type (506) may be a high performance computing workload type. In another example, the workload type (506) may be with a web computing task workload type.

[0053] Fig. 6 is a diagram of an example of an infrastructure controller, according to one example of principles described herein. As will be described below, the infrastructure controller (800) includes a provisioning module (802). The provisioning module (602) detect at least one workload tuned server that matches at least one workload type.

[0054] In the example of Fig. 6, to detect at least one workload tuned server that matches at least one workload type, the provisioning module (602) references a server database (604). In one example, the server database (602) may include information about workload tuned servers (614) such as a Media Access Control (MAC) address (606), RAM size (608), number of processors (610), and workload type (612).

[0055] In this example, the server database (604) includes workload tuned server A (614-1 ). In this example, workload tuned sever A (614-1 ) is associated with MAC address one (606-1 ), RAM size one (608-1 ), number of processors one (610-1 ), and a high performance computing (612-1 ) workload type. When the infrastructure controller (600) receives a provisioning request, the provisioning module (602) reference the server database (604) and matches the provisioning request to a workload tuned server (614) in the server database (604). Fo example, the infrastructure controller (600) receives a provisioning request that uses a high performance computing workload type, the provisioning module (602) reference the server database (604) and matches the provisioning request to workload tuned server A (614-1 ) in the server database (604). The provisioning request is matched to workload tuned server A (614-1 ) because workload tuned server A (614-1 ) has a workload type of high performance computing (612-1 ). [0056] In keeping with the given example, the server database (604) includes workload tuned server B (614-2). In this example, workload tuned sever B (614-2) is associated with MAC address two (606-2), RAM size two (608-2), number of processors two (610-2), and a web computing task (612-2) workload type. When the infrastructure controller (600) receives a provisioning request, the provisioning module (602) reference the server database (604) and matches the provisioning request to a workload tuned server (614) in the server database (604). For example, the infrastructure controller (600) receives a provisioning request that uses a web computing task workload type, the provisioning module (602) reference the server database (604) and matches the provisioning request to workload tuned server B (614-2) in the server database (604). The provisioning request is matched to workload tuned server B (614-2) because workload tuned server B (614-2) has a workload type of a web computing task (612-2).

[0057] While this example has been described with reference to a provisioning request being matched to a workload tuned server based on a workload type, the provisioning request may be matched to the workload tuned server based other criteria. For example, a MAC address, a RAM size, the number of processors on the server, other criteria, or combinations thereof.

[0058] Fig. 7 is a diagram of an example of a data center, according to one example of principles described herein. As will be described below, the data center may include two workload tuned servers. Further, the workload tuned servers may include specific characteristics that define an operation of the workload tuned servers. As a result, one workload type for a computing task may execute more efficiently on one of the workload tuned servers than on another workload tuned server.

[0059] As mentioned above, the data center (700) may include two workload tuned servers. For example, workload tuned server A (702) and workload tuned server B (706). Further, workload tuned server A (702) includes a number of cartridges (704). In one example, the cartridges (704) may be workload type specific. For example, a web computing task workload tuned server such as workload tuned server A (702) has cartridges (704) specific for efficiently executing a web computing task workload type. As a result, workload tuned server A (702) includes a database tuned cartridge (704-1 ), a scripting tuned cartridge (704-2), and a download tuned cartridge (704-3).

[0080] Further, workload tuned server B (706) includes a number of cartridges (708). in one example, the cartridges (708) may be workload specific. Fo example, a high performance computing workload tuned server such as workload tuned server B (706) has cartridges (708) specific for efficiently executing a high performance computing workload type. As a result, workload tuned server B (706) includes a statistics tuned cartridge (708-1 ), a calculus tuned cartridge (708-2), and a linear algebra tuned cartridge (708-3).

[0081] In one example, the linear algebra tuned cartridge (708-3), is used to execute linear algebra equations and may benefit from having a highly parallel processing architecture. In keeping with the given example, the statistics tuned cartridge (708-1 ) may benefit from having a processor with a high clock rate.

[0082] Fig. 8 is a diagram of an example of an optimizing system, according to one example of principles described herein. The optimizing system (800) includes an assigning engine (802), a matching engine (804), a sending engine (808), and a deploying engine (808). In this example, the optimizing system (800) also includes a receiving engine (810), a provisioning engine (812), an integrating engine (814), and a storing engine (816). The engines (802, 804, 806, 808, 810, 812, 814, 816) refer to a combination of hardware and program instructions to perform a designated function. Each of the engines (802, 804, 806, 808, 810, 812, 814, 816) may include a processor and memory. The program instructions are stored in the memory and cause the processor to execute the designated function of the engine.

[0083] The assigning engine (802) assigns at least one workload type to the computing task. As mentioned above, a computing task may have characteristics that make some computer architectures for the more suitable than others computer architectures for processing a computing task. In one example, the characteristics define a workload type that can be used to match the computing task with a workload tuned server. Further, in one example, the assigning engine (802) assigns several workload types to the computing task. In another example, the assigning engine (802) assigns one workload type to the computing task.

[0084] The matching engine (804) matches at least one workload type to at least one workload tuned sever. In one example, the matching engine (804) matches one workload type to at least one workload tuned sever. In another example, the matching engine (804) matches several workload types to at least one workload tuned sever.

[0065] The sending engine (808) sends a provisioning request with at least one workload type to at least one workload tuned server. In one example, the sending engine (808) sends one provisioning request with at least one workload type to at least one workload tuned server. In another example, the sending engine (806) sends several provisioning request with at least one workload type to at least one workload tuned server.

[0068] The deploying engine (808) deploys the computing task on at least one workload tuned server to optimize the computing task. In one example, the deploying engine (808) deploys one computing task on one workload tuned server to optimize the computing task. In another example, the deploying engine (808) deploys several computing tasks on one workload tuned server to optimize the computing task.

[0067] The receiving engine (810) receives a request to deploy a computing task on at least one workload tuned server. In one example, the receiving engine (810) receives one request to deploy a computing task on at least one workload tuned server. In another example, the receiving engine (810) receives several requests to deploy a computing task on at least one workload tuned server.

[0068] The provisioning engine (812) to provision at least one workload tuned server based on the provisioning request. In one example, the provisioning engine (812) provisions at least one workload tuned server based on one provisioning request. In another example, the provisioning engine (812) provisions at least one workload tuned server based on several provisioning requests. [0089] The integrating engine (814) integrates at least one service with at least one workload tuned server. In one example, the integrating engine (814) integrates one service with at least one workload tuned server. In another example, the integrating engine (814) integrates several services with at least one workload tuned server.

[0070] The storing engine (818) stores at least one workload type in a database. Sn one example, the storing engine (816) stores at one workload type in a database. In another example, the storing engine (816) stores at several workload types in a database.

[0071] The preceding description has been presented to illustrate and describe examples of the principles described. This description is not intended to be exhaustive or to limit these principles to any precise form disclosed. Many modifications and variations are possible in light of the above teaching.