CN101741912A - Method, network apparatus and distributed network system for processing computation task - Google Patents
Method, network apparatus and distributed network system for processing computation task Download PDFInfo
- Publication number
- CN101741912A CN101741912A CN200910261811A CN200910261811A CN101741912A CN 101741912 A CN101741912 A CN 101741912A CN 200910261811 A CN200910261811 A CN 200910261811A CN 200910261811 A CN200910261811 A CN 200910261811A CN 101741912 A CN101741912 A CN 101741912A
- Authority
- CN
- China
- Prior art keywords
- grid node
- grid
- calculation task
- task
- node
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Computer And Data Communications (AREA)
- Multi Processors (AREA)
Abstract
The invention provides a method, a network apparatus and a distributed network system for processing computation tasks. The method is applied to the distributed network system, and the grid nodes in the distributed network system interact by using a grid middleware. The method comprises the following steps of: detecting a local load by a first grid node, and transmitting a specific computation task to a second grid node if the local load is detected to exceed a first threshold value; processing the received computation task by the second grid node, and then returning the processed result information to the first grid node. According to the invention, the work pressure of the grid node with heavy load can be diverted, and the idle resources in the distributed network system are adequately utilized.
Description
Technical field
The invention belongs to data communication field, a kind of method, the network equipment and distributed network system (DNS) of handling calculation task is provided.
Background technology
The grid computing middleware is a layer, and it is bundled into the assembly that disperses the supercomputer of a database application.Grid middleware is one of grid key technology.The purpose of research is for the user provides the virtual machine with same DLL (dynamic link library), support the complicated applications problem find the solution with wide area network on the sharing of all kinds of resources.Grid middleware will possess resource dynamic monitoring, masked nodes isomery, optimize functions such as resource selection and collaborative calculating.Grid middleware comprises monitoring resource assembly, calculation services assembly, grid security assembly, fault tolerant service assembly, information service assembly and application schedules assembly.The explanation of successful grid middleware prominent example, its better back-up system provide high-quality service to satisfy the demands for the user.
Under the distributed network equipment environment, the CPU of each equipment does not just have other things except finishing local calculation task can do.Even under extreme condition, for example, to such an extent as to when the very busy edge that reaches collapse of certain appliance services, other CPU can not stretch out " assistance " to it yet.Because commercial network equipment, particularly high-end devices are very high to the requirement of stability,, generally all can remind the attendant to check on the other hand in the mode of reporting to the police when the equipment cpu busy percentage reaches certain threshold value; And this moment, all the other equipment CPU have still is in idle condition, if this CPU can share some calculation tasks, just can reduce the inspection number of times so, has increased the stability of equipment to a certain extent; Meanwhile, also can replace the CPU of high frequency to carry out identical calculation task, objectively reduce the cost expense of device handler with the CPU of low frequency.
Summary of the invention
The technical problem to be solved in the present invention provides a kind of method, the network equipment and distributed network system (DNS) of handling calculation task, with the operating pressure of the network equipment of sharing high capacity, makes full use of resource idle on the distributed network system (DNS).
In order to solve the problems of the technologies described above, the invention provides a kind of method of handling calculation task, be applied in the distributed network system (DNS), the grid node in the described distributed network system (DNS) utilizes grid middleware to carry out it is characterized in that this method comprises alternately:
First grid node detects local load, surpasses first threshold if detect local load, then specific calculation task is sent to second grid node;
Second grid node is handled the described calculation task that receives, and returns result message to first grid node then.
Further, said method has following characteristics: also comprise:
First grid node is collected the resource information of each grid node in the distributed network system (DNS) in good time; Based on described resource information corresponding grid node is carried out formation, set up list of available resources.
Further, said method has following characteristics: described resource information comprises: the CPU usage of the grid node that each is online and memory usage,
Described first grid node detects local load and is specially above first threshold: described first grid node detects local CPU usage and/or memory usage surpasses 60%;
Described second grid node is specially: first grid node is chosen CPU usage and/or memory usage less than 40% grid node from described list of available resources.
Further, said method has following characteristics: described resource information also comprises: the OS Type of the grid node that each is online,
Described first grid node sends to specific calculation task before second grid node, also comprise: whether the OS Type of judging the OS Type of first grid node and second grid node is identical, as if inequality, then the format conversion of described calculation task is become the discernible form of operating system of second grid node.
The present invention also provides a kind of distributed network system (DNS), comprises first grid node and second grid node, and first grid node and second grid node utilize grid middleware to carry out alternately, wherein:
First grid node is used to detect local load, surpasses first threshold if detect local load, then specific calculation task is sent to second grid node;
Second grid node is used to handle the described calculation task that receives, and returns result message to first grid node then.
Further, above-mentioned distributed network system (DNS) has following characteristics:
First grid node, the resource information that also is used for collecting each grid node of distributed network system (DNS); Based on described resource information corresponding grid node is carried out formation, set up list of available resources.
Further, above-mentioned distributed network system (DNS) has following characteristics: described resource information comprises: the CPU usage of the grid node that each is online and memory usage;
First grid node is used to detect local load and is specially above first threshold: detects local CPU usage and/or memory usage and surpasses 60%,
Described second grid node is specially: first grid node is chosen CPU usage and/or memory usage less than 40% grid node from described list of available resources.
Further, above-mentioned distributed network system (DNS) has following characteristics: described resource information also comprises: the OS Type of the grid node that each is online,
First grid node, before being sent to second grid node, specific calculation task also is used for, whether the OS Type of judging first grid node and second grid node is identical, if not, then the format conversion of described calculation task is become the discernible form of operating system of second grid node.
The present invention also provides a kind of network equipment, comprises grid middleware, and the described network equipment utilizes other network equipments in grid middleware and the distributed network system to carry out alternately, wherein,
Described grid middleware is used to detect local load, surpasses first threshold if detect local load, then specific calculation task is sent to destination network device; Be used for the receiving target network equipment and return result corresponding calculation task.
Further, the above-mentioned network equipment has following characteristics: described grid middleware comprises,
The load monitoring module is used to detect local load, surpasses first threshold if detect local load, then notifies the task processing module;
The task processing module, be used to receive described notice after, specific calculation task is sent to destination network device; Be used for the receiving target network equipment and return result corresponding calculation task.
Further, the above-mentioned network equipment has following characteristics: described grid middleware also comprises service management module,
Described load monitoring module, the resource information that also is used for collecting each network equipment of distributed network system (DNS) sends to service management module with the resource information of collecting;
Described service management module, be used to receive described resource information after, based on described resource information corresponding network device is carried out formation, set up list of available resources, described list of available resources is sent to the task processing module;
Described task processing module, also be used to receive the notice that described load monitoring module sends after, the select target network equipment from described list of available resources, specific then calculation task sends to described destination network device.
Further, the above-mentioned network equipment has following characteristics: described resource information comprises: the CPU usage of the network equipment that each is online and memory usage,
Described load monitoring module detects local load and is specially above first threshold: detect local CPU usage and/or memory usage and surpass 60%;
Described destination network device is: described task processing module is chosen CPU usage and/or memory usage less than 40% the network equipment from described list of available resources.
Further, the above-mentioned network equipment has following characteristics: described grid middleware also comprises the heterogeneous platform integrate module, and described resource information also comprises the OS Type of the network equipment that each is online,
Described task processing module is used for specific calculation task is sent to described heterogeneous platform integrate module;
Described heterogeneous platform integrate module, after receiving described calculation task, whether the OS Type of judging local OS Type and described destination network device is identical, if it is inequality, then the format conversion of described calculation task is become the discernible form of operating system of described destination network device, the calculation task after will changing then sends to described destination network device.
To sum up, the method of processing calculation task provided by the invention, the network equipment and distributed network system (DNS), being based on the phenomenon that exists the more wasting of resources in the existing distributed network equipment system proposes, not only can under the prerequisite of not losing the current device performance index, can utilize resource idle in the distributed network system (DNS) significantly, can also strengthen stability, the reduction network equipment cost consumption of the network equipment.
Description of drawings
Fig. 1 is the schematic diagram of the distributed network system (DNS) of the embodiment of the invention;
Fig. 2 is the schematic diagram of the network middleware of the embodiment of the invention;
Fig. 3 is the flow chart of method of the processing calculation task of the embodiment of the invention;
Fig. 4 is the task processing module workflow diagram of the embodiment of the invention.
Embodiment
Be described in further detail below in conjunction with the preferred embodiment of accompanying drawing technical scheme of the present invention.
Fig. 1 is the schematic diagram of the distributed network system (DNS) of the embodiment of the invention, as shown in Figure 1, the distributed network system (DNS) of embodiment utilizes the grid middleware calculation services to integrate and the characteristic of distributing, regard each veneer CPU in the network equipment in the distributed network system (DNS), the particularly network equipment as grid node.
The calculation task of each CPU all can be carried out in the CPU of other any free time in the controlled delivery of grid middleware.When the load of certain grid node in the distributed network equipment system when not being very high, these calculation tasks are still handled on the local hardware of this grid node, when the load on the grid node exceeds predetermined threshold, some particular tasks can be consigned to other idle grid node and handle.
The different grid middleware layers that mainly are between BSP (Board Support Package, plate level support package) layer and the operating system layer of the network equipment of present embodiment with the traditional network equipment.
In the present embodiment distributed network system (DNS), former grid node (also can be considered the former network equipment) and target gridding node (also can be considered destination network device) utilize grid middleware wherein to carry out alternately.
Wherein, former grid node is used to detect local load, surpasses certain threshold level if detect local load, then specific calculation task is sent to the target gridding node;
The target gridding node is used to handle the described calculation task that receives, and returns result to former grid node then.
Can share the operating pressure of the network equipment of high capacity in the distributed network system (DNS) like this, make full use of resource idle on the distributed network system (DNS).
Further, former grid node, the resource information that also is used for collecting each grid node of distributed network system (DNS); Based on described resource information corresponding grid node is carried out formation, set up list of available resources.
Wherein, resource information comprises: the CPU usage of the grid node that each is online and memory usage.Particularly, former grid node detects local CPU usage and/or memory usage surpasses 60%, then specific calculation task is sent to the target gridding node.The target gridding node can be that first grid node is chosen CPU usage and/or memory usage less than 40% grid node from described list of available resources.
Further, resource information also comprises: the OS Type of the grid node that each is online.Further, former grid node is also before sending to second grid node with specific calculation task, whether the OS Type of judging the local operating system and second grid node is identical, if it is inequality, then the format conversion of calculation task is become the discernible form of operating system of second grid node, and then the calculation task after will changing sends to the target gridding node.
Grid middleware in the network equipment of present embodiment specifically is used to detect local load, surpasses certain threshold level if detect local load, then specific calculation task is sent to destination network device; Be used for the result of the receiving target network equipment to corresponding calculation task.
The schematic diagram of the network middleware of present embodiment comprises as shown in Figure 2: load monitoring module and task processing module.The load monitoring module is used to detect local load, surpasses first threshold if detect local load, then notifies the task processing module; The task processing module, be used to receive described notice after, specific calculation task is sent to destination network device, be used for the receiving target network equipment and return result corresponding calculation task.Can share the operating pressure of the network equipment of high capacity in the distributed network system (DNS) like this, make full use of resource idle on the distributed network system (DNS), and then improve the stability of entire environment distributed system operation.
Further, the network middleware of present embodiment also comprises service management module.
Further, the load monitoring module is responsible for collecting the resource information of each grid node in this distributed network system (DNS) in good time, and the resource information of collecting sent to service management module, described resource information comprises: the CPU usage of the grid node that each is online, memory usage and OS Type;
Service management module, after the resource information of each grid node that reception load monitoring module is sent, be responsible for dynamically setting up the list of available resources of grid node formation, and timely replacement is arranged the grid node formation of (for example just sorting by load) with particular order;
The task processing module, form with interrupt service routine exists, when receiving the message of load monitoring module, whether need migration according to local load judgment current task, whether surpass certain threshold level if judge the load of this machine, particularly, judge whether CPU usage and/or memory usage surpass 60%, if surpass, then current some specific task needs migration, then at first inquires about the optimum grid node that is fit to this task from the grid node formation.The selection principle of optimum grid node is: the load of grid node little (for example, CPU usage and/or memory usage are less than 40%), and outer bandwidth free time of band, can guarantee the transmission of task and execution smoothly; By I2C notification target grid node, receiving target gridding node affirmation back package task and sending then.The task processing module can be further divided into task distribution module and receiver module as a result.
Consider the compatibility and the extensibility of distributed network system (DNS), the network middleware of present embodiment may further include the heterogeneous platform integrate module, transform with the interface of being responsible between the different operating system, the format conversion of calculation task is become the discernible form of operating system of second grid node, for example be responsible for the interface conversion between Linux and the Vxworks operating system.
Particularly, the heterogeneous platform integrate module is behind the calculation task that the task processing module of receiving is sent, whether the OS Type of judging local OS Type and described destination network device is identical, if it is inequality, then the format conversion of described calculation task is become the discernible form of operating system of described destination network device, the calculation task after will changing then sends to described destination network device.
Fig. 3 is the flow chart of method of the processing calculation task of the embodiment of the invention, and the method for the processing calculation task of present embodiment is applied in the above-mentioned distributed network system (DNS).As shown in Figure 3, the method for present embodiment comprises following step:
The resource information that S31, former grid node are collected each grid node in the distributed network system (DNS) in good time;
Former grid node carries out formation based on the resource information of collecting to corresponding grid node, sets up list of available resources.
Resource information can comprise: whether grid node is online, OS Type, CPU usage, memory usage etc.
S32, former grid node detect local load;
S33, detect local load and whether surpass certain threshold level, for example, detect local CPU usage and/or memory usage and surpass 60%, if surpass, execution in step S34 then; Otherwise execution in step S36.
S34, former grid node send to the target gridding node with specific calculation task;
The target gridding node is that former grid node is chosen CPU usage and/or memory usage less than 40% grid node from described list of available resources.
Further, before specific calculation task is sent to the target gridding node, former grid node judges earlier whether the OS Type of the OS Type of first grid node and second grid node is identical, if it is inequality, then the format conversion of described calculation task is become the discernible form of operating system of second grid node, and then the calculation task after will changing sends to destination node.
The calculation task that S35, target gridding node processing receive returns result to former grid node then.
S36, handle calculation task in this locality.
Can share the operating pressure of the network equipment of high capacity in the distributed network system (DNS) like this according to the method for the processing calculation task of present embodiment, make full use of resource idle on the distributed network system (DNS), and then improve the stability of entire environment distributed system operation.
Below the workflow of each module in the grid middleware in the network equipment of present embodiment is done detailed explanation.
A, load monitoring module:
In distributed network system (DNS), load monitoring module on the master network device is according to the working condition of each grid node in the distributed environment of its grasp, the resource information that will be related to task run,, OS Type whether online such as grid node, CPU usage, memory usage etc. extract the service management module that consigns to master control and make it generate a list of available resources; Then, the load monitoring module regularly sends to list of available resources by out-band method (being generally High Speed I 2C bus) the load monitoring module of each ply-yarn drill (being other network equipments).
Out-band method is meant in distributed network environment, have a channel that is independent of the business data transmission circuit between each grid node, be specifically designed to the transmission of environment inner control order and sharing of control information with synchronously, just because of these information can not take the business datum bandwidth, so claim that this communication mode is an out-of-band communication.Master control can be discerned each ply-yarn drill by the address.
The load monitoring module of ply-yarn drill receives I2C after the interrupt signal that device produces, obtain and preserve the list of available resources that the load monitoring module on the master network device is sent, decide by checking data then and send which kind of acknowledgement frame to master control (resend message or success message), if do not make a mistake in the transmission course, then send success message to master control; If make a mistake in the transmission course, then send resend message to master control.At last, the notification service administration module upgrades list of available resources.
The load monitoring module of the arbitrary grid node in the distributed network system (DNS) can detect local load, surpasses specific threshold value if detect local load, then notifies the task processing module.By the task processing module current specific calculation task being sent to present load other grid nodes less, that be in idle condition handles.
B, service management module:
After service management module receives the resource information of each grid node that the load monitoring module sends, generate a list of available resources of throwing the net lattice node queue.
Particularly, after receiving resource information, judge whether list of available resources exists, if there is not the list of available resources of grid node formation, then create list of available resources, can insert the list of available resources of creating the grid node formation in proper order according to CPU availability factor, the internal memory availability factor of each node; If had a list of available resources, the resource information of each grid node that extracts according to the load monitoring module is then upgraded the information of corresponding grid node, search then the position again chain go into list of available resources, and after renewal is finished the notification payload monitoring module.
C, task processing module:
The core of task processing module is made up of three aspects, first task encapsulation with unpack, it two is being redirected of the outside uncertain message of task, it three is that execution result is redirected.
Specifically, the task encapsulation is exactly TCB (task control block) information package according to calculation task; It is the TCB information of the TCB package that receives being changed into local corresponding task that task unpacks; Redirected being meant of the outside uncertain message of task sends to task message on the grid node of task actual motion by the out-of-band communication mode, and the task that makes is able to complete operation.Execution result is redirected and is meant in the task run process, and task run finishes the result that the back produces and will be sent on the former grid node by out-of-band communication, so that other tasks relevant with this task are able to trouble-free operation on the former grid node.
Task processing module workflow is as shown in Figure 4:
The performing step of the task processing module of former grid node is:
Whether step S411, the current calculation task of the local load judgment of foundation need migration, if do not need migration then to handle the corresponding calculated task in this locality; If need migration, then continue execution in step S412.
The foundation of migration can be to judge whether this machine load (for example, CPU usage and/or memory usage) surpasses 60%, and the load of the purpose grid node correspondence of migration (for example, CPU usage and/or memory usage) is less than 40%.
Step S412, that the task processing module is selected this grid node according to list of available resources is the most close, and the lightest grid node of load is as the target gridding node of migration.
The TCB information of step S413, encapsulation specific calculation task, change transmission of messages mode is that carry out in the strange land.
Step S414, hang up local task, the notification target grid node starts same task, transmission TCB information.Hang up the local task of needing migration, Qian Yi task will can not continue to carry out in this locality so, no longer take the computational resource of this machine.
Consider the compatibility and the extensibility of distributed network system (DNS), the task processing module of former grid node can also send to the heterogeneous platform integrate module with calculation task information to be migrated.
The performing step of the task processing module of target gridding node is:
After the task processing module of step S421, target gridding node is received the task immigration request of former grid node, start new task, be labeled as the strange land task, then with this task suspension.
The TCB information of step S422, the former grid node of reception is changed into new task TCB information after unpacking, be consistent with the task run step with former grid node.
The scheduled for executing task is hung up in step S423, releasing.
Send it back former grid node with the final message that produces in step S424, the process of will executing the task.
In this strategy, by start the migration that same task realizes task on the target gridding node, the operating system that therefore needs to move on former grid node and the target gridding node need have identical Mission Capability.In addition, on former grid node, be not that all tasks all can realize migration, have only the task of application feature passable, other task of kernel level then can't realize migration.
D, heterogeneous platform integrate module:
The heterogeneous platform integrate module is responsible for the equivalence conversion of TCB under the different operating system platform, particularly, the format conversion of calculation task to be migrated is become the discernible form of operating system of second grid node.For example for Linux and Vxworks, it is exactly the conversion between TCB and the task_struct (task structure body), wherein of paramount importance is the conversion of memory address, by the succession of program and principle of locality as can be known, as long as determined the starting point unanimity of program, in running from now on, also identical execution route can appear so.Therefore need the execution point of the current calculation task of record and the internal memory side-play amount between its starting point, and then change CS (code segment register) and IP content of registers such as (instruction pointer registers), can make former calculation task on the target gridding node, continue operation.
The performing step of the heterogeneous platform integrate module of former grid node:
After the heterogeneous platform integrate module receives the calculation task information of task processing module, determine the OS Type of target gridding node, whether the OS Type of judging the target gridding node is identical with the OS Type of former grid node, if OS Type is identical, then with the TCB information packet of calculation task to be migrated, perhaps other control informations send to the target gridding node; As if the OS Type difference, then the TCB information mapping of origin operation system is converted to the TCB information of purpose operating system, send then.
Can make full use of resource idle on the distributed network system (DNS), and then improve the stability of entire environment distributed system operation with the operating pressure of the network equipment of sharing high capacity according to the network equipment of present embodiment.
Certainly; the present invention also can have other various embodiments; under the situation that does not deviate from spirit of the present invention and essence thereof; those of ordinary skill in the art work as can make various corresponding changes and distortion according to the present invention, but these corresponding changes and distortion all should belong to the protection range of the appended claim of the present invention.
Claims (13)
1. a method of handling calculation task is applied in the distributed network system (DNS), and the grid node in the described distributed network system (DNS) utilizes grid middleware to carry out it is characterized in that this method comprises alternately:
First grid node detects local load, surpasses first threshold if detect local load, then specific calculation task is sent to second grid node;
Second grid node is handled the described calculation task that receives, and returns result message to first grid node then.
2. the method for claim 1 is characterized in that: also comprise:
First grid node is collected the resource information of each grid node in the distributed network system (DNS) in good time; Based on described resource information corresponding grid node is carried out formation, set up list of available resources.
3. method as claimed in claim 2 is characterized in that: described resource information comprises: the CPU usage of the grid node that each is online and memory usage,
Described first grid node detects local load and is specially above first threshold: described first grid node detects local CPU usage and/or memory usage surpasses 60%;
Described second grid node is specially: first grid node is chosen CPU usage and/or memory usage less than 40% grid node from described list of available resources.
4. as claim 2 or 3 described methods, it is characterized in that: described resource information also comprises: the OS Type of the grid node that each is online,
Described first grid node sends to specific calculation task before second grid node, also comprise: whether the OS Type of judging the OS Type of first grid node and second grid node is identical, as if inequality, then the format conversion of described calculation task is become the discernible form of operating system of second grid node.
5. a distributed network system (DNS) comprises first grid node and second grid node, and first grid node and second grid node utilize grid middleware to carry out it is characterized in that alternately:
First grid node is used to detect local load, surpasses first threshold if detect local load, then specific calculation task is sent to second grid node;
Second grid node is used to handle the described calculation task that receives, and returns result message to first grid node then.
6. distributed network system (DNS) as claimed in claim 5 is characterized in that:
First grid node, the resource information that also is used for collecting each grid node of distributed network system (DNS); Based on described resource information corresponding grid node is carried out formation, set up list of available resources.
7. distributed network system (DNS) as claimed in claim 6 is characterized in that: described resource information comprises: the CPU usage of the grid node that each is online and memory usage;
First grid node is used to detect local load and is specially above first threshold: detects local CPU usage and/or memory usage and surpasses 60%,
Described second grid node is specially: first grid node is chosen CPU usage and/or memory usage less than 40% grid node from described list of available resources.
8. as claim 6 or 7 described distributed network system (DNS)s, it is characterized in that: described resource information also comprises: the OS Type of the grid node that each is online,
First grid node, before being sent to second grid node, specific calculation task also is used for, whether the OS Type of judging first grid node and second grid node is identical, if not, then the format conversion of described calculation task is become the discernible form of operating system of second grid node.
9. a network equipment comprises grid middleware, and the described network equipment utilizes other network equipments in grid middleware and the distributed network system to carry out it is characterized in that alternately:
Described grid middleware is used to detect local load, surpasses first threshold if detect local load, then specific calculation task is sent to destination network device; Be used for the receiving target network equipment and return result corresponding calculation task.
10. the network equipment as claimed in claim 9 is characterized in that: described grid middleware comprises,
The load monitoring module is used to detect local load, surpasses first threshold if detect local load, then notifies the task processing module;
The task processing module, be used to receive described notice after, specific calculation task is sent to destination network device; Be used for the receiving target network equipment and return result corresponding calculation task.
11. the network equipment as claimed in claim 10 is characterized in that: described grid middleware also comprises service management module,
Described load monitoring module, the resource information that also is used for collecting each network equipment of distributed network system (DNS) sends to service management module with the resource information of collecting;
Described service management module, be used to receive described resource information after, based on described resource information corresponding network device is carried out formation, set up list of available resources, described list of available resources is sent to the task processing module;
Described task processing module, also be used to receive the notice that described load monitoring module sends after, the select target network equipment from described list of available resources, specific then calculation task sends to described destination network device.
12. the network equipment as claimed in claim 11 is characterized in that: described resource information comprises: the CPU usage of the network equipment that each is online and memory usage,
Described load monitoring module detects local load and is specially above first threshold: detect local CPU usage and/or memory usage and surpass 60%;
Described destination network device is: described task processing module is chosen CPU usage and/or memory usage less than 40% the network equipment from described list of available resources.
13. as the claim 11 or the 12 described network equipments, it is characterized in that: described grid middleware also comprises the heterogeneous platform integrate module, described resource information also comprises the OS Type of the network equipment that each is online,
Described task processing module is used for specific calculation task is sent to described heterogeneous platform integrate module;
Described heterogeneous platform integrate module, after receiving described calculation task, whether the OS Type of judging local OS Type and described destination network device is identical, if it is inequality, then the format conversion of described calculation task is become the discernible form of operating system of described destination network device, the calculation task after will changing then sends to described destination network device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910261811A CN101741912A (en) | 2009-12-30 | 2009-12-30 | Method, network apparatus and distributed network system for processing computation task |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910261811A CN101741912A (en) | 2009-12-30 | 2009-12-30 | Method, network apparatus and distributed network system for processing computation task |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101741912A true CN101741912A (en) | 2010-06-16 |
Family
ID=42464795
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910261811A Pending CN101741912A (en) | 2009-12-30 | 2009-12-30 | Method, network apparatus and distributed network system for processing computation task |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101741912A (en) |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102111300A (en) * | 2011-03-14 | 2011-06-29 | 浪潮(北京)电子信息产业有限公司 | Monitoring system and method for guaranteeing business continuity |
| CN102117226A (en) * | 2011-03-18 | 2011-07-06 | 浪潮(北京)电子信息产业有限公司 | Resource dispatching system and resource dispatching method |
| CN102170474A (en) * | 2011-04-22 | 2011-08-31 | 广州杰赛科技股份有限公司 | Method and system for dynamic scheduling of virtual resources in cloud computing network |
| CN102420751A (en) * | 2011-11-28 | 2012-04-18 | 曙光信息产业(北京)有限公司 | Load judgment system and method for report sampling system |
| CN102419718A (en) * | 2011-10-28 | 2012-04-18 | 浪潮(北京)电子信息产业有限公司 | Resource scheduling method |
| CN102761832A (en) * | 2011-04-28 | 2012-10-31 | 中国移动通信集团河南有限公司 | Message distribution method and device |
| CN103078764A (en) * | 2012-12-28 | 2013-05-01 | 国家计算机网络与信息安全管理中心 | Operational monitoring system and method based on virtual computing task |
| CN103248526A (en) * | 2012-02-08 | 2013-08-14 | 迈普通信技术股份有限公司 | Communication equipment and method for achieving out-of-band monitoring and management, and master-slave switching method |
| CN103369524A (en) * | 2013-07-30 | 2013-10-23 | 东莞宇龙通信科技有限公司 | Terminal and data processing method |
| CN104579593A (en) * | 2013-10-23 | 2015-04-29 | 鸿富锦精密工业（深圳）有限公司 | Data monitoring system and method |
| CN104717153A (en) * | 2013-12-12 | 2015-06-17 | 国际商业机器公司 | Resource over-subscription |
| WO2016000365A1 (en) * | 2014-06-30 | 2016-01-07 | 中兴通讯股份有限公司 | Method and apparatus for path computation |
| WO2016091003A1 (en) * | 2014-12-08 | 2016-06-16 | 中兴通讯股份有限公司 | Method for implementing service collaborative scheduling, computing board, and storage medium |
| WO2016123972A1 (en) * | 2015-02-02 | 2016-08-11 | 中兴通讯股份有限公司 | Load balancing method and load balancing apparatus |
| CN106233276A (en) * | 2014-03-14 | 2016-12-14 | 亚马逊科技公司 | The coordination access control of network-accessible block storage device |
| CN106844055A (en) * | 2017-01-25 | 2017-06-13 | 北京百分点信息科技有限公司 | A kind of execution method and apparatus of task |
| CN103870322B (en) * | 2012-12-17 | 2017-06-30 | 联发科技股份有限公司 | Method, non-transitory computer-readable medium, the heterogeneous multi-core system of control task transfer |
| CN107295110A (en) * | 2017-08-16 | 2017-10-24 | 网宿科技股份有限公司 | Processing method, fringe node, service server and the system of calculating task |
| CN107577561A (en) * | 2017-09-14 | 2018-01-12 | 郑州云海信息技术有限公司 | A kind of data reconstruction method, device, equipment and computer-readable recording medium |
| CN108092971A (en) * | 2017-12-13 | 2018-05-29 | 新华三信息安全技术有限公司 | A kind of method and device of processing business message |
| CN109815003A (en) * | 2017-11-21 | 2019-05-28 | 中国移动通信有限公司研究院 | Cooperated computing method, system, smart machine and medium between smart machine |
| CN109876446A (en) * | 2019-01-14 | 2019-06-14 | 珠海金山网络游戏科技有限公司 | A kind of method and apparatus for detecting game textures and grid and occupying distribution in memory |
| CN109933420A (en) * | 2019-04-02 | 2019-06-25 | 深圳市网心科技有限公司 | Node tasks dispatching method, electronic equipment and system |
| CN110111458A (en) * | 2019-04-16 | 2019-08-09 | 深圳市多度科技有限公司 | Realize method, system and door access machine that face is deployed to ensure effective monitoring and control of illegal activities |
| WO2020001161A1 (en) * | 2018-06-29 | 2020-01-02 | 华为技术有限公司 | Task processing method and apparatus |
| CN112019581A (en) * | 2019-05-30 | 2020-12-01 | 华为技术有限公司 | Method and device for scheduling task processing entities |
| CN113051957A (en) * | 2019-12-26 | 2021-06-29 | 浙江宇视科技有限公司 | Data analysis method and system |
| CN115051997A (en) * | 2022-07-12 | 2022-09-13 | 中国电信股份有限公司 | Method and device for scheduling distributed resources, storage medium and electronic equipment |
-
2009
- 2009-12-30 CN CN200910261811A patent/CN101741912A/en active Pending
Cited By (41)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102111300A (en) * | 2011-03-14 | 2011-06-29 | 浪潮(北京)电子信息产业有限公司 | Monitoring system and method for guaranteeing business continuity |
| CN102117226A (en) * | 2011-03-18 | 2011-07-06 | 浪潮(北京)电子信息产业有限公司 | Resource dispatching system and resource dispatching method |
| CN102170474A (en) * | 2011-04-22 | 2011-08-31 | 广州杰赛科技股份有限公司 | Method and system for dynamic scheduling of virtual resources in cloud computing network |
| CN102761832A (en) * | 2011-04-28 | 2012-10-31 | 中国移动通信集团河南有限公司 | Message distribution method and device |
| CN102419718A (en) * | 2011-10-28 | 2012-04-18 | 浪潮(北京)电子信息产业有限公司 | Resource scheduling method |
| CN102420751A (en) * | 2011-11-28 | 2012-04-18 | 曙光信息产业(北京)有限公司 | Load judgment system and method for report sampling system |
| CN103248526B (en) * | 2012-02-08 | 2016-06-08 | 迈普通信技术股份有限公司 | Realize signal equipment, the method for out-of-band supervision control management |
| CN103248526A (en) * | 2012-02-08 | 2013-08-14 | 迈普通信技术股份有限公司 | Communication equipment and method for achieving out-of-band monitoring and management, and master-slave switching method |
| CN103870322B (en) * | 2012-12-17 | 2017-06-30 | 联发科技股份有限公司 | Method, non-transitory computer-readable medium, the heterogeneous multi-core system of control task transfer |
| CN103078764A (en) * | 2012-12-28 | 2013-05-01 | 国家计算机网络与信息安全管理中心 | Operational monitoring system and method based on virtual computing task |
| CN103369524A (en) * | 2013-07-30 | 2013-10-23 | 东莞宇龙通信科技有限公司 | Terminal and data processing method |
| CN104579593A (en) * | 2013-10-23 | 2015-04-29 | 鸿富锦精密工业（深圳）有限公司 | Data monitoring system and method |
| CN104579593B (en) * | 2013-10-23 | 2018-01-16 | 鸿富锦精密工业（深圳）有限公司 | Data monitoring system and method |
| CN104717153A (en) * | 2013-12-12 | 2015-06-17 | 国际商业机器公司 | Resource over-subscription |
| CN104717153B (en) * | 2013-12-12 | 2018-12-25 | 国际商业机器公司 | Computer implemented method and system for management work load in a network |
| CN106233276A (en) * | 2014-03-14 | 2016-12-14 | 亚马逊科技公司 | The coordination access control of network-accessible block storage device |
| CN106233276B (en) * | 2014-03-14 | 2019-05-14 | 亚马逊科技公司 | The coordination admission control of network-accessible block storage device |
| CN105282023A (en) * | 2014-06-30 | 2016-01-27 | 中兴通讯股份有限公司 | Method and device for calculating path |
| WO2016000365A1 (en) * | 2014-06-30 | 2016-01-07 | 中兴通讯股份有限公司 | Method and apparatus for path computation |
| WO2016091003A1 (en) * | 2014-12-08 | 2016-06-16 | 中兴通讯股份有限公司 | Method for implementing service collaborative scheduling, computing board, and storage medium |
| CN105743802B (en) * | 2014-12-08 | 2020-02-14 | 中兴通讯股份有限公司 | Method for realizing service cooperative scheduling and computing single board |
| CN105743802A (en) * | 2014-12-08 | 2016-07-06 | 中兴通讯股份有限公司 | Method of realizing business cooperative scheduling, and calculating single board |
| WO2016123972A1 (en) * | 2015-02-02 | 2016-08-11 | 中兴通讯股份有限公司 | Load balancing method and load balancing apparatus |
| CN105991458A (en) * | 2015-02-02 | 2016-10-05 | 中兴通讯股份有限公司 | Load balancing method and load balancing device |
| CN106844055B (en) * | 2017-01-25 | 2020-02-28 | 北京百分点信息科技有限公司 | Task execution method and device |
| CN106844055A (en) * | 2017-01-25 | 2017-06-13 | 北京百分点信息科技有限公司 | A kind of execution method and apparatus of task |
| CN107295110B (en) * | 2017-08-16 | 2020-08-04 | 网宿科技股份有限公司 | Processing method of computing task, edge node, service server and system |
| CN107295110A (en) * | 2017-08-16 | 2017-10-24 | 网宿科技股份有限公司 | Processing method, fringe node, service server and the system of calculating task |
| CN107577561A (en) * | 2017-09-14 | 2018-01-12 | 郑州云海信息技术有限公司 | A kind of data reconstruction method, device, equipment and computer-readable recording medium |
| CN109815003A (en) * | 2017-11-21 | 2019-05-28 | 中国移动通信有限公司研究院 | Cooperated computing method, system, smart machine and medium between smart machine |
| CN109815003B (en) * | 2017-11-21 | 2020-12-18 | 中国移动通信有限公司研究院 | Cooperative computing method and system between intelligent devices, intelligent device and medium |
| CN108092971A (en) * | 2017-12-13 | 2018-05-29 | 新华三信息安全技术有限公司 | A kind of method and device of processing business message |
| WO2020001161A1 (en) * | 2018-06-29 | 2020-01-02 | 华为技术有限公司 | Task processing method and apparatus |
| CN109876446A (en) * | 2019-01-14 | 2019-06-14 | 珠海金山网络游戏科技有限公司 | A kind of method and apparatus for detecting game textures and grid and occupying distribution in memory |
| CN109933420A (en) * | 2019-04-02 | 2019-06-25 | 深圳市网心科技有限公司 | Node tasks dispatching method, electronic equipment and system |
| CN110111458A (en) * | 2019-04-16 | 2019-08-09 | 深圳市多度科技有限公司 | Realize method, system and door access machine that face is deployed to ensure effective monitoring and control of illegal activities |
| CN112019581A (en) * | 2019-05-30 | 2020-12-01 | 华为技术有限公司 | Method and device for scheduling task processing entities |
| CN112019581B (en) * | 2019-05-30 | 2022-02-25 | 华为技术有限公司 | Method and device for scheduling task processing entities |
| CN113051957A (en) * | 2019-12-26 | 2021-06-29 | 浙江宇视科技有限公司 | Data analysis method and system |
| CN115051997A (en) * | 2022-07-12 | 2022-09-13 | 中国电信股份有限公司 | Method and device for scheduling distributed resources, storage medium and electronic equipment |
| CN115051997B (en) * | 2022-07-12 | 2024-03-19 | 中国电信股份有限公司 | Distributed resource scheduling method and device, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101741912A (en) | Method, network apparatus and distributed network system for processing computation task | |
| CN106330769B (en) | Service processing method and server | |
| US10223140B2 (en) | System and method for network function virtualization resource management | |
| US8032780B2 (en) | Virtualization based high availability cluster system and method for managing failure in virtualization based high availability cluster system | |
| CN103210619B (en) | For nothing lock and the zero-copy messaging plan of communication network application | |
| WO2019001092A1 (en) | Load balancing engine, client, distributed computing system, and load balancing method | |
| CN108781229A (en) | Control the method and server of the reorientation of MEC applications | |
| CN107241281B (en) | Data processing method and device | |
| CN103227838B (en) | A kind of multi-load equilibrium treatment apparatus and method | |
| CN105007337A (en) | Cluster system load balancing method and system thereof | |
| CN101902395A (en) | The system and method for the resource on the management host | |
| CN100517221C (en) | Efficient multiprocessor system and methods thereof | |
| US9104488B2 (en) | Support server for redirecting task results to a wake-up server | |
| CN103154897A (en) | Core abstraction layer for telecommunication network applications | |
| CN101873005A (en) | Method for realizing balanced acquisition of electric energy | |
| CN108600355A (en) | A kind of mist computational load balance dispatching system and method | |
| WO2010027609A2 (en) | Load balancing for services | |
| EP4270204A1 (en) | Multi-cloud interface adaptation method and system based on micro-service, and storage medium | |
| CN107977254B (en) | Method for responding to request in cloud data system and computer-readable storage medium | |
| CN104426694A (en) | Method and device for adjusting virtual-machine resources | |
| CN111124640A (en) | Task allocation method and system, storage medium and electronic device | |
| CN107220102A (en) | The dormancy of virtual machine and awakening method and system | |
| CN103677983A (en) | Scheduling method and device of application | |
| US9594596B2 (en) | Dynamically tuning server placement | |
| CN112565390B (en) | Service distribution method, device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20100616 |