CN108270603A - A kind of dispatching method and management system - Google Patents

Abstract

The embodiment of the present invention provides a kind of dispatching method and management system, for solving the technical issues of scheduling of resource ability of resource management system in the prior art is poor.This method includes：It determines that there are what user carried out to operate with the first of the relevant operation plan of current data task for setting；The first operation of response, is shown and the relevant at least one metric of current data task by display interface；Wherein, at least one metric include user setting for the resource requirement and restrictive condition of current data task in the process of implementation, resource requirement is used to indicate the requirement to the hardware resource of server, and restrictive condition is used to indicate the service requirement to server；Based at least one metric, operation plan corresponding with current data task is determined；Scheduling of resource is carried out based on operation plan, determines at least one destination server corresponding with current data task, at least one destination server disclosure satisfy that resource requirement and restrictive condition.

Description

A scheduling method and management system

technical field

The present invention relates to the field of communication technology, in particular to a scheduling method and a management system.

Background technique

In the current big data ecosystem, the requirements are becoming more and more complex, and the traditional Map-Reduce (a programming model) computing task is no longer suitable for all requirements. Therefore, in Hadoop 2.0 (a distributed system infrastructure) or In YARN (a new Hadoop resource management system), the type of computing tasks can be defined by the user. In Spark (an open source cluster computing environment similar to Hadoop), the flow of computing tasks is more flexible. In addition, as the cluster size continues to grow, heterogeneous resources are used to run big data workloads, including open power, x86, SolarisSparc, and more. In the current resource management environment (such as YARN, MESOS, etc.), big data computing tasks may have various requirements in terms of runtime resources, environment and task constraints.

But at present, big data resource management software (such as YARN and Mesos, etc.) that are mainly used cannot schedule computing tasks in a more flexible mode. Although YARN can support multiple dimensions (such as CPU, memory, hard disk, etc.), it cannot Support for more complex requirements. For example, it cannot support more scheduling constraints (such as operating system type, etc.) and runtime scheduling constraints (such as "run this task on the last host"), etc., and its scheduler's configurable items It is fixed and unchanged. Users cannot expand or customize the scheduling constraints in the real production environment. The scheduling configuration process is inconvenient. It may not be possible to determine the most suitable server to perform tasks during the resource scheduling process, and even cause damage to the server. resource consumption.

To sum up, it can be known that the resource management system in the prior art has a technical problem of poor resource scheduling capability.

Contents of the invention

Embodiments of the present invention provide a scheduling method and a management system, which are used to solve the technical problem of poor resource scheduling capability of the resource management system in the prior art.

In the first aspect, the implementation of the present invention provides a scheduling method, including:

Determine that there is a first operation performed by the user for setting a scheduling plan related to the current data task;

In response to the first operation, at least one metric information related to the current data task is displayed through a display interface; wherein the at least one metric information includes resource requirements set by the user during execution of the current data task and a restriction condition, the resource requirement is used to indicate the requirement on the hardware resources of the server, and the restriction condition is used to indicate the operation requirement on the server;

Determine a scheduling plan corresponding to the current data task based on the at least one metric information;

Resource scheduling is performed based on the scheduling plan, and at least one target server corresponding to the current data task is determined, and the at least one target server can meet the resource requirement and the restriction condition.

Optionally, before determining that there is a user operation performed by a user in the display interface of the management system for setting a scheduling plan related to the current data task, the method further includes:

Acquiring newly added metric information of the user for the newly added resource metric of the resource metric collector in the management system, the resource metric collector being used to collect multiple resource metrics of the corresponding server and the metric parameters corresponding to each resource metric;

And, obtaining the restrictive condition information set by the user through the restrictive condition component in the management system for the operation of the server that executes the data task;

In response to the first operation, displaying at least one metric information related to the current data task through a display interface, including:

In response to the first operation, display the at least one metric information on the display interface based on the newly added metric information and the restriction condition information.

Optionally, the determining the scheduling plan corresponding to the current data task based on the at least one metric information includes:

Determine whether there is a second operation performed on the at least one metric information in the display interface, where the second operation is an operation for adjusting a metric parameter corresponding to the at least one metric information;

If the second operation exists, generate a scheduling plan corresponding to the current data task based on the adjusted metric parameter and the at least one metric information, or, if the second operation does not exist, based on the at least one metric The information and its corresponding measurement parameters generate a scheduling plan corresponding to the current data task.

Optionally, performing resource scheduling based on the scheduling plan, and determining at least one target server corresponding to the current data task includes:

Determine at least one scheduling instruction corresponding to the scheduling plan, and each scheduling instruction in the at least one scheduling instruction indicates corresponding resource requirements and constraints on the server;

Resource scheduling is performed based on the at least one scheduling instruction, and at least one target server corresponding to the current data task is determined.

Optionally, the determining at least one target server corresponding to the current data task includes:

Determine at least one target server among the multiple servers connected to the management system that can meet the resource requirements and constraints corresponding to the scheduling plan.

In a second aspect, an embodiment of the present invention provides a management system, including:

A first determining module, configured to determine that there is a first operation performed by a user for setting a scheduling plan related to the current data task;

An operation module, configured to display at least one metric information related to the current data task through a display interface in response to the first operation; wherein, the at least one metric information includes the execution time for the current data task set by the user Resource requirements and constraints in the process, the resource requirements are used to indicate the requirements for the hardware resources of the server, and the constraints are used to indicate the operating requirements for the server;

A second determining module, configured to determine a scheduling plan corresponding to the current data task based on the at least one metric information;

A scheduling module, configured to perform resource scheduling based on the scheduling plan, determine at least one target server corresponding to the current data task, and the at least one target server can meet the resource requirement and the restriction condition.

Optionally, the system also includes:

The obtaining module is used to obtain the user's newly added resource measurement collector in the management system before determining that there is a user operation for setting a scheduling plan related to the current data task in the display interface of the management system. newly added metric information of resource metric, the resource metric collector is used to collect multiple resource metrics of the corresponding server and the metric parameters corresponding to each resource metric; Information about restrictions on the operation of servers performing data tasks;

The operation module is configured to: display the at least one metric information on the display interface based on the newly added metric information and the restriction condition information in response to the first operation.

Optionally, the second determination module is used for:

Determine whether there is a second operation performed on the at least one metric information in the display interface, where the second operation is an operation for adjusting a metric parameter corresponding to the at least one metric information;

If the second operation exists, generate a scheduling plan corresponding to the current data task based on the adjusted metric parameter and the at least one metric information, or, if the second operation does not exist, based on the at least one metric The information and its corresponding measurement parameters generate a scheduling plan corresponding to the current data task.

Optionally, the scheduling module is used for:

Determine at least one scheduling instruction corresponding to the scheduling plan, and each scheduling instruction in the at least one scheduling instruction indicates corresponding resource requirements and constraints on the server;

Resource scheduling is performed based on the at least one scheduling instruction, and at least one target server corresponding to the current data task is determined.

Optionally, the scheduling module is used for:

Determine at least one target server among the multiple servers connected to the management system that can meet the resource requirements and constraints corresponding to the scheduling plan.

In the embodiment of the present invention, when it is determined that there is a first operation performed by the user for setting the scheduling plan related to the current data task, the management system can respond to the first operation and display at least one metric related to the scheduling plan through the display unit Information, at least one piece of metric information may include resource requirements and constraints during the execution of the current data task, the resource requirements may be used to indicate the limitations of the hardware resources of the server for executing the data task, and the constraints are used to indicate the execution of the current data task The operation requirements of the task on the server, and then, based on one less measurement information, the scheduling plan corresponding to the current data task can be determined, and resource scheduling can be performed based on the scheduling plan, and at least one target server corresponding to the current data task can be determined. A target server can meet the resource requirements and constraints. Therefore, during the scheduling process, displaying at least one measurement information can facilitate the user to view the configuration process of the scheduling plan, improve the accuracy and user experience during the configuration process, and at the same time, the displayed measurement information Including the resource requirements and constraints set by the user, so that the scheduling plan is more in line with the actual needs of the user in the process of processing data, making the scheduling of server resources more accurate, and through user customization, the management system can meet the needs of diversification Computing task requirements, realize fine-grained resource scheduling, and improve the management effect of resources.

Description of drawings

Fig. 1 is a schematic diagram of the architecture of the management system in the embodiment of the present invention;

Fig. 2 is the main flowchart of the scheduling method in the embodiment of the present invention;

FIG. 3 is a schematic diagram of newly added resource metrics for setting a scheduling plan in an embodiment of the present invention;

FIG. 4 is a schematic diagram of setting constraints of a scheduling plan in an embodiment of the present invention;

FIG. 5 is a schematic diagram of a visualized graphical interface of the management system in an embodiment of the present invention;

6 is a schematic diagram of a visualized graphical interface of the management system in an embodiment of the present invention;

Fig. 7 is a block diagram of the management system in the embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

First, the system architecture applied in the embodiment of the present invention is introduced.

In the embodiment of the present invention, the management system may be a system for resource management of server clusters, which may assign corresponding servers to perform task processing based on user requests for data tasks. In practical applications, the management system may be a distributed management system composed of a master station (dispatch center) and multiple sub-stations, and each sub-station may be a corresponding server.

As shown in FIG. 1 , it is a schematic diagram of the architecture of the management system and its corresponding application environment in the embodiment of the present invention. The management system may mainly include a scheduling engine and a server cluster. Among them, the scheduling engine can perform corresponding calculations on the tasks requested by the user based on the scheduling plan created by the user, such as batch processing tasks for a large amount of data, and allocate corresponding server resources to perform tasks based on the scheduling plan. The server cluster mainly It is used for scheduling by the scheduling engine to process various tasks.

Specifically, the management system in the embodiment of the present invention may include four components, namely: a pluggable resource metric collector, a pluggable job/task constraint component, a graphical interface for configuring the scheduling plan, and translating the scheduling plan parser into resource requirements.

Among them, the pluggable resource metric collector can be deployed in each server of the server cluster, and is responsible for collecting the resource metric corresponding to the server, and the resource metric can be used to indicate the hardware resource and its attributes of the corresponding server, such as CPU core , the size of the memory, etc.

The pluggable job/task constraint component can be used to provide a corresponding interface for users to add personalized job/task constraints, and realize runtime scheduling constraints.

The graphical interface can be used to manage the scheduling plan, and the resource configuration related to the scheduling plan can be displayed through the graphical interface, such as resource metrics and constraints. Users can create/modify/view scheduling plans on a visual graphical interface.

The scheduling plan parser is located in the scheduling engine, and is used to parse and translate the scheduling plan into corresponding scheduling instructions, and then the scheduler will select the most appropriate resource container to allocate to this task to achieve fine-grained resource scheduling.

In addition, the term "and/or" in this article is only an association relationship describing associated objects, which means that there may be three relationships, for example, A and/or B, which may mean: A exists alone, A and B exist at the same time, There are three cases of B alone. In addition, the character "/" in this article generally indicates that the contextual objects are an "or" relationship.

Preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings.

As shown in FIG. 2, an embodiment of the present invention provides a scheduling method, which is applied to a management system. The management system can be used in a system for scheduling and managing server resources. The method includes the following steps:

S11: Determine that there is a first operation performed by the user for setting a scheduling plan related to the current data task;

S12: In response to the first operation, display at least one measurement information related to the scheduling plan through the display interface; wherein, the at least one measurement information includes resource requirements and constraints set by the user during the execution of the current data task, and the resource requirements are used Used to indicate the requirements for the hardware resources of the server, and the constraints are used to indicate the operating requirements for the server;

S13: Based on at least one metric information, determine a scheduling plan corresponding to the current data task;

S14: Perform resource scheduling based on the scheduling plan, determine at least one target server corresponding to the current data task, and at least one target server can meet resource requirements and constraints.

In the embodiment of the present invention, when it is determined that there is a first operation performed by the user for setting the scheduling plan of the current data task, at least one metric information related to the scheduling plan can be displayed on the display interface by responding to the first operation, which includes For information such as resource requirements and constraints during the execution of the current data task, the user can clearly view or set the measurement information based on the display interface, and then based on the measurement information, the scheduling plan corresponding to the current data task can be generated, then Based on the resource scheduling in the chassis of the scheduling plan, at least one target server suitable for executing the current data task can be determined, so that the accuracy of the scheduling process is high.

In practical applications, the first operation may be an operation performed by the user to call out an entry or page capable of setting a schedule related to a data task. For example, the first operation may be performed by a designer through an input device in the management system (such as mouse) or an operating body (such as a finger) to click on the corresponding options in the display interface, or it can also be an operation performed in other ways, such as an input operation by inputting a corresponding operation instruction, etc. The examples of the present invention are This is not specifically limited.

The current data task can be a large amount of data requested by the user to the server through the terminal device for processing, and it can also be network data, such as WeChat, music, and video used in the communication and social process, or group buying, e-mail, etc. Businesses, mobile payments, etc. will continuously generate massive amounts of information data, and of course it can also be other big data that needs to be processed. Those skilled in the art can set the data tasks to be processed according to actual needs, which is not specifically limited in the embodiment of the present invention.

Optionally, before S11, the user can set the metrics corresponding to the scheduling plan of the data task through the corresponding components included in the management system, and the management system can obtain the corresponding information set by the user. The process may include: determining the newly added metric information of the user for the newly added resource metric of the resource metric collector in the management system, the resource metric collector is used to collect multiple resource metrics of the corresponding server and the metric parameters corresponding to each resource metric, and , to determine the restrictive condition information set by the user through the restrictive condition component in the management system for the operation of the server executing the data task.

Wherein, the metric parameter may be an operational parameter capable of characterizing the corresponding resource metric, such as a version parameter, a rate parameter, a storage capacity parameter, a kernel parameter, and the like. The operation restriction may refer to the restriction on the operation of the host in processing the data task, for example, it is allowed to execute in multiple or one host.

Specifically, during the setting process, it may include but not limited to the following settings.

Setting 1: The user can collect multiple fine-grained resource metrics of the corresponding server through the resource metric collector component contained in the management system, while the open source component YARN used in the prior art can only support several fixed dimensions (such as CPU , memory, hard disk, etc.), and multiple fine-grained resource metrics that can be collected by the resource metric collector in the management system in the implementation of the present invention can include GNU Compiler Collection (GNU Compiler Collection, GCC), Java Operating Environment (Java Runtime Environment, JRE), software usage authorization (License), etc.

For example, as shown in Figure 3, users can implement the AgentInterface interface to define different implementation classes, that is, add new metrics collection plug-ins or add new metrics to the default Resource Metric Agents configuration file to define more fine-grained resource metrics collect. For example, in Figure 3, by adding a new <metric/> node in the Resource Metric Agents configuration file, the default JRE version metric collection program will run the script "java-version" to obtain the Java version and report it to the cluster scheduling engine. More <metric/> nodes can be added, such as GCC Version (GCC version), License Name (license item), etc., to realize the collection of metrics for multiple resources.

Setting 2: Users can add personalized job/task restriction conditions through the pluggable job/task restriction component (ie, restriction component) to realize runtime scheduling restriction.

For example, as shown in Figure 4, the user can implement the Agent Interface to define a new implementation class or add a new <metric/> node in the configuration file of the Default Contraint Agent to increase the dependency restriction to define complex task dependency semantics, such as "Taskof this job will be run on same hosts", which requires "all tasks of this job will be run on the same hosts", and so on.

In the embodiment of the present invention, after the first operation is detected, by responding to the first operation, at least one metric information related to the scheduling plan can be displayed in the corresponding visual display interface through the graphical interface component included in the management system, that is, the scheduling Corresponding resource requirements and constraints can be displayed on a graphical interface, that is, a user interface (User Interface, UI), which is referred to as a display interface in this paper.

As shown in FIG. 5 , it is a schematic diagram of a visualized graphical interface in the management system. Users can create/modify/view the scheduling plan on the graphical interface, and the user implements the scheduling plug-in interface through the aforementioned settings 1 and 2 to add resource requirements and constraints, and these new requirements and constraints will be added in the UI, such as What is shown in the dotted box 1 in FIG. 5 is the newly added resource metric, and what is shown in the dotted box 2 is the corresponding restriction condition.

In practical applications, users can also set the corresponding information in the graphical interface,

Therefore, in the embodiment of the present invention, pluggable and extensible resource dimensions and job/task restrictions can be realized through multiple components included in the management system, and the corresponding measurement information is displayed through a visual graphical interface, so that users can perform corresponding The management and viewing avoid the drawbacks in the prior art that the corresponding scheduling plan information cannot be viewed and/or modified during the scheduling process, and the visual display interface improves user experience.

In the embodiment of the present invention, when determining the scheduling plan corresponding to the current data task based on at least one metric information, it may be determined whether there is a second operation for at least one metric information in the display interface, and the second operation is for adjusting An operation on a metric parameter corresponding to at least one metric information.

That is, when at least one metric information is displayed on the display interface, it may be detected whether there is a setting operation by the user on the resource metric and the restriction condition on the display interface. For example, the user can set the measurement requirements of the License Name in the resource measurement (that is, configure the measurement parameters) through the display interface shown in Figure 4, such as modifying the authorization of the scope of use of the software from "matlab" to "Java", or changing " The metric requirement corresponding to GCC Version" is changed from ">=2.6" to "=2.6", and so on.

Or, the user can also set the restriction conditions, for example, the option "Task running on same job'stask hosts" under the option column "runtime constraint conditions (ie, runtime constraints)" in Figure 5 will be selected by " "Required" is changed to "Preferred", etc.

After selecting the preferred option, the user can name and save the scheduling plan. Specifically, if there is a second operation, generate a scheduling plan corresponding to the current data task based on the adjusted measurement parameter and at least one measurement information, or , if there is no second operation, generate a scheduling plan corresponding to the current data task based on at least one metric information and corresponding metric parameters.

Of course, in practical applications, users can save multiple versions of scheduling plans through settings, and different scheduling plans can have different names and different measurement parameters, etc., and the scheduling plans can be called directly later.

Therefore, in the embodiment of the present invention, since the graphical interface of the corresponding measurement information displayed in the management system is also an operable interface, the configuration of the relevant measurement parameters of the measurement information by the user can also help to improve the accuracy of the scheduling plan and improve the scheduling efficiency. Flexibility during plan configuration.

Optionally, after the scheduling engine obtains the generated scheduling plan, it can use the parser component in the scheduling engine to translate the plan, that is, set 1 and 2 resource measurement requirements and run-time job/task constraints into scheduling instructions, and then schedule The scheduler will select the most suitable resource container to allocate to this task, that is, the scheduler can perform resource scheduling based on the scheduling plan, and determine at least one target server corresponding to the current data task, that is, the current data task can be executed through the at least one target server , to achieve fine-grained resource scheduling.

In practical applications, the at least one target server may be a server in the server cluster in the management system, which can meet the resource requirements and constraints corresponding to the scheduling plan, wherein the number of at least one server may be related to the constraints in the scheduling plan , for example, when the restriction condition is "run the task on the same host", there may be only one target server corresponding to the task assigned to process the current data.

The process of scheduling and managing resources by the management system in the embodiment of the present invention will be described below in conjunction with Flowchart 6 .

First, get users to define resource metrics collected from computing nodes through pluggable resource metric collectors, and define job/task constraints through pluggable job/task constraint components;

Then, when the detection user creates a job scheduling plan through the graphical interface, the aforementioned defined items (ie resource independence and limiting conditions) will be displayed in the graphical interface as candidate conditions;

Furthermore, once a schedule is created, it can be bound to a job. Users can bind a schedule to a job through job configuration parameters. For example, for a map reduce job, specify conf.set ("SchedulingPlan", <PLAN_NAME>);

Then, the parser component in the scheduling engine will parse the resource metric requirements and runtime job/task constraints into scheduler-readable instructions;

Finally, the scheduling engine in the management system schedules jobs/tasks according to scheduling instructions.

Of course, after the execution is completed, the management system can also reclaim the resources, or, during the execution process, the pluggable resource metric collector can be used to collect the corresponding resource metrics of the server, such as temporary memory usage and so on.

Therefore, in the embodiment of the present invention, by providing a pluggable interface for the management system to define the metrics of resources and limit rules for runtime computing tasks, and setting a graphical interface to create a scheduling plan for jobs, and a parser to convert the scheduling The plan is translated into the resource requirements of the resource scheduling engine, and the resource scheduling engine will allocate resources according to the resource requirements of each application, thereby improving the management system's management of resources.

As shown in FIG. 7 , based on the same inventive concept, an embodiment of the present invention also provides a management system. The system architecture of the management system is shown in FIG. 1 . The management system may include a first determination module 201 , an operation module 202 , a second determination module 203 and a scheduling module 204 .

Wherein, the first determining module 201 may be configured to determine that there is a first operation performed by the user for setting a scheduling plan related to the current data task;

The operation module 202 may be configured to display at least one metric information related to the current data task through a display interface in response to the first operation; Resource requirements and constraints during the execution process, the resource requirements are used to indicate the requirements for the hardware resources of the server, and the constraints are used to indicate the operating requirements for the server;

The second determining module 203 may be configured to determine a scheduling plan corresponding to the current data task based on the at least one metric information;

The scheduling module 204 may be configured to perform resource scheduling based on the scheduling plan, determine at least one target server corresponding to the current data task, and the at least one target server can meet the resource requirement and the restriction condition.

Optionally, the system may further include an obtaining module, which may be used to obtain the user's information on the user's information about the current data task before determining that there is a user operation for setting a scheduling plan related to the current data task in the display interface of the management system. new metric information of newly added resource metrics of the resource metric collector in the management system, the resource metric collector is used to collect multiple resource metrics of the corresponding server and the metric parameters corresponding to each resource metric; Restriction information set by the restriction component in the management system for the operation of the server that executes the data task;

The operation module 202 may be configured to display the at least one metric information on the display interface based on the newly added metric information and the restriction condition information in response to the first operation.

Optionally, the second determining module 203 may be configured to: determine whether there is a second operation performed on the at least one metric information in the display interface, and the second operation is to adjust the at least one metric information corresponding to The operation of the metric parameter; if the second operation exists, generate a scheduling plan corresponding to the current data task based on the adjusted metric parameter and the at least one metric information, or, if the second operation does not exist, A scheduling plan corresponding to the current data task is generated based on the at least one metric information and corresponding metric parameters.

Optionally, the scheduling module 204 may be configured to determine at least one scheduling instruction corresponding to the scheduling plan, and each scheduling instruction in the at least one scheduling instruction indicates corresponding resource requirements and constraints on the server; furthermore, based on the At least one scheduling instruction performs resource scheduling, and determines at least one target server corresponding to the current data task.

Optionally, the scheduling module 204 may also be configured to determine at least one target server among multiple servers connected to the management system that can meet the resource requirements and constraints corresponding to the scheduling plan.

In the embodiment of the present invention, when it is determined that there is a first operation performed by the user for setting the scheduling plan related to the current data task, the management system can respond to the first operation and display at least one metric related to the scheduling plan through the display unit Information, at least one piece of metric information may include resource requirements and constraints during the execution of the current data task, the resource requirements may be used to indicate the limitations of the hardware resources of the server for executing the data task, and the constraints are used to indicate the execution of the current data task The operation requirements of the task on the server, and then, based on one less measurement information, the scheduling plan corresponding to the current data task can be determined, and resource scheduling can be performed based on the scheduling plan, and at least one target server corresponding to the current data task can be determined. A target server can meet the resource requirements and constraints. Therefore, during the scheduling process, displaying at least one measurement information can facilitate the user to view the configuration process of the scheduling plan, improve the accuracy and user experience during the configuration process, and at the same time, the displayed measurement information Including the resource requirements and constraints set by the user, so that the scheduling plan is more in line with the actual needs of the user in the process of processing data, making the scheduling of server resources more accurate, and through user customization, the management system can meet the needs of diversification Computing task requirements, realize fine-grained resource scheduling, and improve the management effect of resources.

The process and examples of the scheduling method in the aforementioned embodiments of Figures 2-6 are also applicable to the management system of this embodiment. Through the foregoing detailed description of the scheduling method, those skilled in the art can clearly understand the implementation of the management system in this embodiment For example, for the sake of brevity, it will not be described in detail here.

Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, systems, or computer program products. Accordingly, the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims (10)

1. a kind of dispatching method, applied to management system, which is characterized in that including：

It determines that there are what user carried out to operate with the first of the relevant operation plan of current data task for setting；

First operation is responded, is shown by display interface and believed with the relevant at least one measurement of the current data task Breath；Wherein, at least one metric include user setting for the current data task in the process of implementation Resource requirement and restrictive condition, the resource requirement is used to indicate the requirement to the hardware resource of server, the limitation item Part is used to indicate the service requirement to server；

Based at least one metric, operation plan corresponding with the current data task is determined；

Scheduling of resource is carried out based on the operation plan, determines at least one destination service corresponding with the current data task Device, at least one destination server disclosure satisfy that the resource requirement and the restrictive condition.

2. the method as described in claim 1, which is characterized in that there are users in the display interface for determining the management system Before what is carried out is used to set the user's operation with the relevant operation plan of current data task, the method further includes：

Newly-increased metric of the user for the newly-increased resource measurement of resource measurement collector in the management system is obtained, it is described Resource measurement collector is used to collect the multinomial resource measurement metric parameter corresponding with each resource measurement of respective server；

And obtain the service for execution data task that the user is set by restrictive condition component in the management system The restrictive condition information of device operation；

Response first operation, is shown and the relevant at least one measurement of the current data task by display interface Information, including：

First operation is responded, based on the newly-increased metric and the restrictive condition information, in the display interface Show at least one metric.

3. method as claimed in claim 2, which is characterized in that described based at least one metric, determining and institute The corresponding operation plan of current data task is stated, including：

It determines whether there is and is operated for the second of metric progress at least one described in the display interface, described second It operates as adjusting the operation of the corresponding metric parameter of at least one metric；

If there are the described second operation, worked as based on the metric parameter after adjustment and at least one metric generation with described If the corresponding operation plan of preceding data task or, there is no the described second operation, based at least one metric and its Corresponding metric parameter generates operation plan corresponding with the current data task.

4. the method as described in any claims of claim 1-3, which is characterized in that resource tune is carried out based on the operation plan Degree determines at least one destination server corresponding with the current data task, including：

Determine the corresponding at least one dispatch command of the operation plan, each dispatch command at least one dispatch command Indicate corresponding resource requirement and restrictive condition to server；

Scheduling of resource is carried out based at least one dispatch command, is determined corresponding at least one with the current data task Destination server.

5. method as claimed in claim 4, which is characterized in that described to determine corresponding with the current data task at least one A destination server, including：

It determines to disclosure satisfy that the corresponding resource requirement of the operation plan and limit in multiple servers of the management system connection At least one destination server of condition processed.

6. a kind of management system, which is characterized in that including：

First determining module to be used to set and the relevant operation plan of current data task for determining that there are what user carried out First operation；

Operation module for responding first operation, is shown relevant extremely with the current data task by display interface A few metric；Wherein, at least one metric include user setting for the current data task Resource requirement and restrictive condition in the process of implementation, the resource requirement, which is used to indicate, wants the hardware resource of server It asks, the restrictive condition is used to indicate the service requirement to server；

Second determining module for being based at least one metric, determines tune corresponding with the current data task Degree plan；

Scheduler module carries out scheduling of resource for being based on the operation plan, determines corresponding extremely with the current data task A few destination server, at least one destination server disclosure satisfy that the resource requirement and the restrictive condition.

7. system as claimed in claim 6, which is characterized in that the system also includes：

Acquisition module to be used for setting and current number for there are what user carried out in the display interface for determining the management system Before user's operation according to the relevant operation plan of task, user is obtained for resource measurement collector in the management system The newly-increased metric of newly-increased resource measurement, the resource measurement collector are used to collect the multinomial resource measurement of respective server Metric parameter corresponding with each resource measurement；And it obtains the user and is set by restrictive condition component in the management system The restrictive condition information for the server operation for performing data task put；

The operation module is used for：First operation is responded, based on the newly-increased metric and the restrictive condition information, At least one metric is shown in the display interface.

8. system as claimed in claim 7, which is characterized in that second determining module is used for：

It determines whether there is and is operated for the second of metric progress at least one described in the display interface, described second It operates as adjusting the operation of the corresponding metric parameter of at least one metric；

If there are the described second operation, worked as based on the metric parameter after adjustment and at least one metric generation with described If the corresponding operation plan of preceding data task or, there is no the described second operation, based at least one metric and its Corresponding metric parameter generates operation plan corresponding with the current data task.

9. the system as described in any claims of claim 6-8, which is characterized in that the scheduler module is used for：

Determine the corresponding at least one dispatch command of the operation plan, each dispatch command at least one dispatch command Indicate corresponding resource requirement and restrictive condition to server；

Scheduling of resource is carried out based at least one dispatch command, is determined corresponding at least one with the current data task Destination server.

10. system as claimed in claim 9, which is characterized in that the scheduler module is used for：

It determines to disclosure satisfy that the corresponding resource requirement of the operation plan and limit in multiple servers of the management system connection At least one destination server of condition processed.