CN103336714A - Operation scheduling method and device - Google Patents
Operation scheduling method and device Download PDFInfo
- Publication number
- CN103336714A CN103336714A CN2013102475903A CN201310247590A CN103336714A CN 103336714 A CN103336714 A CN 103336714A CN 2013102475903 A CN2013102475903 A CN 2013102475903A CN 201310247590 A CN201310247590 A CN 201310247590A CN 103336714 A CN103336714 A CN 103336714A
- Authority
- CN
- China
- Prior art keywords
- group
- priority
- weights
- time
- submitted
- 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
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an operation scheduling method and a device. The method comprises the following steps: partitioning the resources of a cloud computing platform into N groups, and distributing submitted work into corresponding groups according to specifications given during submission, wherein N is a natural number; utilizing resources of corresponding groups during running of each job; specific to each group, if the resources of the group meet the requirement of all jobs submitted to the group, directly scheduling all the jobs in the group; and if the resources of the group cannot meet the requirement of all the jobs submitted to the group, putting the jobs submitted to the group into a queue for waiting, and scheduling the jobs according to the priority and waiting time of each job in the group. Due to the adoption of the technical scheme of the invention, the situation that low-priority businesses are not scheduled when jobs of high priorities are submitted continuously can be avoided.
Description
Technical field
The present invention relates to technical field of the computer network, be specifically related to a kind of job scheduling method and device.
Background technology
Cloud computing is a kind of computation schema that dynamic telescopic virtualized resource is provided with service manner by computer network.Cloud computing platform provides the cloud platform of storage and calculation services.
The calculating that the resource of utilizing cloud computing platform to provide is finished is commonly referred to operation.The kind of operation can be divided into that concurrent program does not have communication type and concurrent program has communication type.
Do not have communication type for concurrent program, each subtask of this operation can at any time start, and does not need to start simultaneously with other subtasks, because need not communicate by letter between the subtask, the calculating that oneself is handled in each subtask just can directly finish.Therefore the cloud computing platform scheduler can be dispatched different subtasks at different time, and the scheduling mode of handling this dispatching requirement is called the task level scheduling.
Therefore for concurrent program communication type is arranged, each subtask of this operation needs mutual communication, needs etc. could to begin to calculate after all subtasks all start.After also namely needing the current resource that has all subtasks that can start this operation of cloud computing platform, could be with this job initiation, otherwise, need wait until that other operations are calculated to finish, discharge enough resources, could start this operation.The scheduling mode of handling this dispatching requirement is called the job class scheduling.
Maui is a scheduler of increasing income that extensively is used, and multiple schedulable condition can be set, the scheduler of Chang Zuowei cloud computing platform.Be to handle the calculation task that needs the job class scheduling based on the cloud computing platform of Maui.The most frequently used job class dispatching method of this platform is the packet-weighted method of method of first-in, first-out and band priority.Because method of first-in, first-out is very simple, relative merits are also known, therefore do not give unnecessary details here.Following emphasis is told about the packet-weighted method of band priority:
1. determine grouping number N, and every group weights W i, wherein i ∈ { 1 ... N};
2. the resource division with cloud computing platform becomes the N group, and every group of shared resource ratio equals the ratio of every group weight.Every group of maximum spendable resource must not surpass the value that calculates according to its ratio;
3. need specify during submit job to operate in which group, and the resource that this operation needs can not surpass the total resources of this group, can failure otherwise submit to;
4. specify the priority of this operation when submitting computational tasks to simultaneously, generally be divided into basic, normal, high three priority;
5. the operation in same group, when sufficient resources was arranged, which priority no matter operation be, all can be by management and running immediately;
6. if this group resource is inadequate, then job queue's situation can take place, the priority of operation plays a role at this moment.When idling-resource, operation high priority operation earlier, secondly medium priority is low priority at last.Queuing time scheduling is successively afterwards earlier pressed in the operation that priority is identical.
There is following shortcoming in the packet-weighted dispatching method of existing this band priority: in same group, occur under the situation of queuing, if the priority ratio high operation of the job priority of lowest priority of having lined up in the tabulation is ceaselessly arranged to be submitted to, the operation of these lowest priorities will be died of hunger so, namely can not get the chance of carrying out forever.
Summary of the invention
In view of the above problems, the present invention has been proposed in order to a kind of a kind of job scheduling method and device that overcomes the problems referred to above or address the above problem at least in part is provided.
According to one aspect of the present invention, a kind of job scheduling method is provided, this method comprises:
The resource division of cloud computing platform is become the N group, with the operation submitted to the assignment of allocation when submitting in the group of correspondence; Wherein N is natural number; Each operation utilizes the resource of corresponding group when operation;
For each group,
If the resource of this group can satisfy the demand of the whole operations that are submitted to this group, then directly dispatch the whole operations in this group;
If the resource of this group can not satisfy the demand of the whole operations that are submitted to this group, the operation that then will be submitted to this group is put into the formation of this group and is waited in line, and according to the priority of each operation in this group with wait in line the time and dispatch.
Alternatively, described priority according to each operation in this group is dispatched with the time of waiting in line and is comprised:
For each operation in this group arranges the priority weights, priority is more high, and the priority weights are more big;
For each operation in this group arranges the waiting time weights, waiting time is more long, waiting time weights more big;
The priority weights and waiting time weights and be the comprehensive weights of operation;
The comprehensive big operation of weights during priority scheduling should be organized.
Alternatively, this method further comprises: for each group,
Estimate the operation that can not postponed the working time of the next operation that need dispatch of this group working time, then this operation of priority scheduling if exist during this forms a team to be listed as;
Wherein, operation estimates the working time of portable band when operation is submitted to.
Alternatively, this method further comprises: estimate working time if surpassed it working time of an operation, then finish the operation of this operation.
Alternatively, this method further comprises: for periodic real time job, reserve resource requirement at its periodic execution time window for this operation;
If this periodic real time job is not activated in its periodic execution time window, then no longer reserve resource requirement for it.
According to another aspect of the present invention, a kind of job scheduling device is provided, this device comprises: grouped element and press the weights scheduling unit;
Described grouped element, the resource division that is suitable for cloud computing platform becomes the N group, with the operation submitted to the assignment of allocation when submitting in the group of correspondence; Wherein N is natural number; Each operation utilizes the resource of corresponding group when operation;
The described weights scheduling unit of pressing is suitable for each group if the resource of this group can satisfy the demand of the whole operations that are submitted to this group, is then directly dispatched the whole operations in this group; If the resource of this group can not satisfy the demand of the whole operations that are submitted to this group, the operation that then will be submitted to this group is put into the formation of this group and is waited in line, and according to the priority of each operation in this group with wait in line the time and dispatch.
Alternatively, the described weights scheduling unit of pressing, be suitable for each group, for each operation in this group arranges the priority weights, priority is more high, the priority weights are more big, for each operation in this group arranges the waiting time weights, waiting time is more long, waiting time weights more big, and with the priority weights and waiting time weights and be the comprehensive weights of operation, and priority scheduling should group in the comprehensive big operation of weights.
Alternatively, this device further comprises: by the estimated time scheduling unit;
Described by the estimated time scheduling unit, be suitable for for each group, if this exist in forming a team to be listed as estimate working time can be with should group next operation that needs postponement working time of the operation that dispatch, then this operation of priority scheduling; Wherein, operation estimates the working time of portable band when operation is submitted to.
Alternatively, described by the estimated time scheduling unit, be suitable for when having surpassed it and estimate working time the working time of an operation, finishing the operation of this operation.
Alternatively, this device further comprises: press the periodic scheduling unit;
The described periodic scheduling unit of pressing, be suitable for periodic real time job, reserve resource requirement at its periodic execution time window for this operation, and, if this periodic real time job is not activated in its periodic execution time window, then no longer reserve resource requirement for it.
This resource division with cloud computing platform according to the present invention becomes N group, and in the group of correspondence, wherein N is natural number with the operation submitted to the assignment of allocation when submitting to, and each operation utilizes the resource of corresponding group when operation; For each group, if the resource of this group can satisfy the demand of the whole operations that are submitted to this group, then directly dispatch the whole operations in this group; If the resource of this group can not satisfy the demand of the whole operations that are submitted to this group, the operation that then will be submitted to this group is put into the formation of this group and is waited in line, and the technical scheme of dispatching according to the priority of each operation in this group and the time of waiting in line, because during the operation in every group of wait queue of scheduling, not only considered the factor of priority, and consideration can be waited in line the factor of time, therefore, can avoid only considering in the scheme of priority, in same group, occur under the situation of queuing, if ceaselessly there is the priority ratio high operation of the job priority of lowest priority of having lined up in the tabulation to submit to, the operation of these lowest priorities can not get carrying out the problem of chance so.
Above-mentioned explanation only is the general introduction of technical solution of the present invention, for can clearer understanding technological means of the present invention, and can be implemented according to the content of instructions, and for above and other objects of the present invention, feature and advantage can be become apparent, below especially exemplified by the specific embodiment of the present invention.
Description of drawings
By reading hereinafter detailed description of the preferred embodiment, various other advantage and benefits will become cheer and bright for those of ordinary skills.Accompanying drawing only is used for the purpose of preferred implementation is shown, and does not think limitation of the present invention.And in whole accompanying drawing, represent identical parts with identical reference symbol.In the accompanying drawings:
Fig. 1 shows a kind of process flow diagram of job scheduling method according to an embodiment of the invention;
Fig. 2 shows a kind of according to an embodiment of the invention structural drawing of job scheduling device;
Fig. 3 shows a kind of according to an embodiment of the invention structural drawing of job scheduling device.
Embodiment
Exemplary embodiment of the present disclosure is described below with reference to accompanying drawings in more detail.Though shown exemplary embodiment of the present disclosure in the accompanying drawing, yet should be appreciated that and to realize the disclosure and the embodiment that should do not set forth limits here with various forms.On the contrary, it is in order to understand the disclosure more thoroughly that these embodiment are provided, and can with the scope of the present disclosure complete convey to those skilled in the art.
Fig. 1 shows a kind of process flow diagram of job scheduling method according to an embodiment of the invention.As shown in Figure 1, this method comprises:
Step S110 becomes the N group with the resource division of cloud computing platform, with the operation submitted to the assignment of allocation when submitting in the group of correspondence; Wherein N is natural number; Each operation utilizes the resource of corresponding group when operation.
For example, establishing cloud computing platform has 1000 nodes (i.e. 1000 computing equipments), is divided into four groups, is specially: A group, 300 nodes; The B group, 200 nodes; The C group, 100 nodes; The D group, 400 nodes.
Need specify during submit job to operate in which group, can formulate the strategy of which kind of job run in which group according to actual conditions, not do restriction in the present invention.Which group operation is assigned to, and then utilizes the resource operation of this group, and for example certain operation is assigned to the A group, and then this operation need utilize the resource operation in 300 nodes in the A group.
Step S120 for each group, if the resource of this group can satisfy the demand of the whole operations that are submitted to this group, then directly dispatches the whole operations in this group.
Be example with the A group, if the current operation that is submitted to the A group amounts to the resource that needs 260 nodes, and the A group has the resource of 300 nodes, satisfies the demands fully, and the therefore directly current whole operations that are submitted to the A group of scheduling move these operations.
Step S130, for each group, if the resource of this group can not satisfy the demand of the whole operations that are submitted to this group, the operation that then will be submitted to this group is put into the formation of this group and is waited in line, and according to the priority of each operation in this group with wait in line the time and dispatch.
Be example with the A group still, if if the current operation that is submitted to the A group of current submission amounts to the resource that needs 500 nodes, and the A group has only the resource of 300 nodes, can not satisfy whole demands, therefore these operations need be put into wait queue, according to certain strategy, successively dispatch.In an embodiment of the present invention, according to the priority of each operation in the A group with wait in line the time and dispatch.
In the method shown in Figure 1, because during the operation in every group of wait queue of scheduling, not only considered the factor of priority, and consideration can be waited in line the factor of time, therefore, can avoid only considering in the scheme of priority, in same group, occur under the situation of queuing, if ceaselessly there is the priority ratio high operation of the job priority of lowest priority of having lined up in the tabulation to submit to, the operation of these lowest priorities can not get carrying out the problem of chance so.
In one embodiment of the invention, the priority according to each operation in this group in the method shown in Figure 1 is dispatched with the time of waiting in line and is specifically as follows: for each operation in this group arranges the priority weights, priority is more high, and the priority weights are more big; For each operation in this group arranges the waiting time weights, waiting time is more long, waiting time weights more big; The priority weights and waiting time weights and be the comprehensive weights of operation; The comprehensive big operation of weights during priority scheduling should be organized.
Being divided into basic, normal, high 3 kinds of priority-level with the priority with operation is example, 3 kinds of priority mapping are become 3 kinds of numerical value of priority weights, priority is more high, the priority weights are more big: the priority weights of lower priority job are 0, the priority weights of medium priority operation are 1000, and the priority weights of high priority operation are 4000.
In addition, time of waiting in line of operation also is mapped to the waiting time weights, for example shines upon by description, namely lined up 10 seconds, then weights are 10 waiting time, line up 100 seconds, and then weights are 100 waiting time, by that analogy.
Then for each operation in the formation, with its priority weights and waiting time weights add up as the comprehensive weights of this job scheduling, comprehensive weights are more high, more can preferentially obtain scheduling.
Be according to priority to dispatch so generally, but considered queuing time simultaneously, prevent that lower priority job is hungry to death.By above-mentioned example as can be known, lower priority job waited at most 4000 seconds, will come new high priority operation front.And the medium priority operation waited at most 3000 seconds, will come new high priority operation front.Guaranteed so not occur same operation being arranged forever in the situation of tail of the queue.
Also there is following shortcoming in the packet-weighted dispatching method of existing this band priority: under same group of situation that occurs lining up, if the resource of this group can not satisfy first operation that comes team's head, even having the littler operation of resource requirement to organize residual resource at this, the back carries out, this little operation can not be scheduled yet, and causes the handling capacity of cloud computing platform lower.
To this, among the present invention on the basis of scheme shown in Figure 1, further provided following solution: for each group, if this exist in forming a team to be listed as estimate working time can be with should group next operation that needs postponement working time of the operation that dispatch, then this operation of priority scheduling.Wherein, operation estimates the working time of portable band when operation is submitted to.
For example, organizing with the above-mentioned C that mentions is example, have the resource of 100 nodes in this C group, if the current operation of having dispatched and having moved has taken the resource of 80 nodes, the next operation that need dispatch (operation that ranked first the position in the formation) need be used the resource of 30 nodes, and have the resource of 10 nodes to be released after the fastest 2 hours in occupied 80 nodes, that is to say that the next operation that needs to dispatch can only just have enough resources to be scheduled after 2 hours.According to existing scheme, can only wait until the operation that ranked first the position of going after 2 hours in the scheduling queue this moment, idle 20 node resources just can only free in these two hours.And according to technical scheme of the present invention, if exist resource requirement to be less than or equal to 20 node resources in the formation and estimate working time less than 2 hours operation, then can these operations of priority scheduling, because the scheduling to these little operations can't influence operation plan originally, thereby the resource of taking full advantage of.
In one embodiment of the invention, the accuracy of estimating working time that provides when impelling the user to guarantee submit job is being taked following scheme: estimate working time if surpassed it working time of an operation, then finish the operation of this operation.Namely surpassed its estimated time the working time of an operation, then this operation is just stopped automatically by platform.
Do not influence the scheme of little operation of the scheduling time of other operations in this priority scheduling queue queue, improved the throughput of platform.
Also there is following shortcoming in the packet-weighted dispatching method of existing this band priority: if there is real-time to require very high important operation, need periodically fixed point operation, this dispatching method can not guarantee can move this operation so, because the group at this operation place may resource remain seldom at once.
To this, on the basis of scheme shown in Figure 1, provided following solution again among the present invention: for periodic real time job, reserve resource requirement at its periodic execution time window for this operation; If this periodic real time job is not activated in its periodic execution time window, then no longer reserve resource requirement for it.
Namely according to the solution of the present invention, for periodic real time job, reserve resource requirement at its periodic window of carrying out, these resources are only used for this periodic real time job, can identify the reserved resource of this operation with job name.Because understand some deviation the start-up time of periodic real time job, therefore in an embodiment of the present invention, its periodic execution in the window, when if this operation is not activated, and other job initiations, and when platform did not have other resources to come for these other operations uses, these other operations need enter queue queue, and can not use reserved resource.If this periodic real time job is not activated in its periodic execution time window, then no longer reserve resource requirement for it.
This to periodic real time job, reserve the scheme of its resource requirement, guaranteed the real-time requirement of important operation.
Fig. 2 shows a kind of according to an embodiment of the invention structural drawing of job scheduling device.As shown in Figure 2, this business dispatching device 200 comprises: grouped element 201 and press weights scheduling unit 202.
Grouped element 201, the resource division that is suitable for cloud computing platform becomes the N group, with the operation submitted to the assignment of allocation when submitting in the group of correspondence; Wherein N is natural number; Each operation utilizes the resource of corresponding group when operation.
For example, establishing cloud computing platform has 1000 nodes (i.e. 1000 computing equipments), is divided into four groups, is specially: A group, 300 nodes; The B group, 200 nodes; The C group, 100 nodes; The D group, 400 nodes.
Need specify during submit job to operate in which group, can formulate the strategy of which kind of job run in which group according to actual conditions, not do restriction in the present invention.Which group operation is assigned to, and then utilizes the resource operation of this group, and for example certain operation is assigned to the A group, and then this operation need utilize the resource operation in 300 nodes in the A group.
Press weights scheduling unit 202, be suitable for each group if the resource of this group can satisfy the demand of the whole operations that are submitted to this group, is then directly dispatched the whole operations in this group; If the resource of this group can not satisfy the demand of the whole operations that are submitted to this group, the operation that then will be submitted to this group is put into the formation of this group and is waited in line, and according to the priority of each operation in this group with wait in line the time and dispatch.
In the device shown in Figure 2, because during the operation in every group of wait queue of scheduling, not only considered the factor of priority, and consideration can be waited in line the factor of time, therefore, can avoid only considering in the scheme of priority, in same group, occur under the situation of queuing, if ceaselessly there is the priority ratio high operation of the job priority of lowest priority of having lined up in the tabulation to submit to, the operation of these lowest priorities can not get carrying out the problem of chance so
Fig. 3 shows a kind of according to an embodiment of the invention structural drawing of job scheduling device.As shown in Figure 3, this business dispatching device 300 comprises: grouped element 301, press weights scheduling unit 302, by estimated time scheduling unit 303 and press periodic scheduling unit 304.
Grouped element 301, the resource division that is suitable for cloud computing platform becomes the N group, with the operation submitted to the assignment of allocation when submitting in the group of correspondence; Wherein N is natural number; Each operation utilizes the resource of corresponding group when operation.
Press weights scheduling unit 302, be suitable for each group if the resource of this group can satisfy the demand of the whole operations that are submitted to this group, is then directly dispatched the whole operations in this group; If the resource of this group can not satisfy the demand of the whole operations that are submitted to this group, the operation that then will be submitted to this group is put into the formation of this group and is waited in line, and according to the priority of each operation in this group with wait in line the time and dispatch.
In one embodiment of the invention, press weights scheduling unit 302, be suitable for each group, for each operation in this group arranges the priority weights, priority is more high, the priority weights are more big, for each operation in this group arranges the waiting time weights, waiting time is more long, waiting time weights more big, and with the priority weights and waiting time weights and be the comprehensive weights of operation, and priority scheduling should group in the comprehensive big operation of weights.
Being divided into basic, normal, high 3 kinds of priority-level with the priority with operation is example, 3 kinds of priority mapping are become 3 kinds of numerical value of priority weights, priority is more high, the priority weights are more big: the priority weights of lower priority job are 0, the priority weights of medium priority operation are 1000, and the priority weights of high priority operation are 4000.In addition, time of waiting in line of operation also is mapped to the waiting time weights, for example shines upon by description, namely lined up 10 seconds, then weights are 10 waiting time, line up 100 seconds, and then weights are 100 waiting time, by that analogy.Then for each operation in the formation, with its priority weights and waiting time weights add up as the comprehensive weights of this job scheduling, comprehensive weights are more high, more can preferentially obtain scheduling.
Be according to priority to dispatch so generally, but considered queuing time simultaneously, prevent that lower priority job is hungry to death.By above-mentioned example as can be known, lower priority job waited at most 4000 seconds, will come new high priority operation front.And the medium priority operation waited at most 3000 seconds, will come new high priority operation front.Guaranteed so not occur same operation being arranged forever in the situation of tail of the queue
In one embodiment of the invention, by estimated time scheduling unit 303, be suitable for for each group, if this exist in forming a team to be listed as estimate working time can be with should group next operation that needs postponement working time of the operation that dispatch, then this operation of priority scheduling; Wherein, operation estimates the working time of portable band when operation is submitted to.
For example, organizing with the above-mentioned C that mentions is example, have the resource of 100 nodes in this C group, if the current operation of having dispatched and having moved has taken the resource of 80 nodes, the next operation that need dispatch (operation that ranked first the position in the formation) need be used the resource of 30 nodes, and have the resource of 10 nodes to be released after the fastest 2 hours in occupied 80 nodes, that is to say that the next operation that needs to dispatch can only just have enough resources to be scheduled after 2 hours.According to existing scheme, can only wait until the operation that ranked first the position of going after 2 hours in the scheduling queue this moment, in these two hours idle 20 node resources just in vain at leisure.And according to technical scheme of the present invention, if exist resource requirement to be less than or equal to 20 node resources in the formation and estimate working time less than 2 hours operation, then can these operations of priority scheduling, thus the resource of taking full advantage of.
The accuracy of estimating working time that provides when impelling the user to guarantee submit job is suitable for finishing the operation of this operation when having surpassed it and estimate working time the working time of an operation by estimated time scheduling unit 303.
In one embodiment of the invention, press periodic scheduling unit 304, be suitable for periodic real time job, reserve resource requirement at its periodic execution time window for this operation, and, if this periodic real time job is not activated in its periodic execution time window, then no longer reserve resource requirement for it.
Namely for periodic real time job, press periodic scheduling unit 304, reserve resource requirement at its periodic window of carrying out, these resources are only used for this periodic real time job, can identify the reserved resource of this operation with job name.Because understand some deviation the start-up time of periodic real time job, therefore in an embodiment of the present invention, its periodic execution in the window, when if this operation is not activated, and other job initiations, and when platform did not have other resources to come for these other operations uses, these other operations need enter queue queue, and can not use reserved resource.If this periodic real time job is not activated in its periodic execution time window, then no longer reserve resource requirement for it.
Need to prove:
Intrinsic not relevant with any certain computer, virtual system or miscellaneous equipment with demonstration at this algorithm that provides.Various general-purpose systems also can be with using based on the teaching at this.According to top description, it is apparent constructing the desired structure of this type systematic.In addition, the present invention is not also at any certain programmed language.Should be understood that and to utilize various programming languages to realize content of the present invention described here, and the top description that language-specific is done is in order to disclose preferred forms of the present invention.
In the instructions that provides herein, a large amount of details have been described.Yet, can understand, embodiments of the invention can be put into practice under the situation of these details not having.In some instances, be not shown specifically known method, structure and technology, so that not fuzzy understanding of this description.
Similarly, be to be understood that, in order to simplify the disclosure and to help to understand one or more in each inventive aspect, in the description to exemplary embodiment of the present invention, each feature of the present invention is grouped together in single embodiment, figure or the description to it sometimes in the above.Yet the method for the disclosure should be construed to the following intention of reflection: namely the present invention for required protection requires the more feature of feature clearly put down in writing than institute in each claim.Or rather, as following claims reflected, inventive aspect was to be less than all features of the disclosed single embodiment in front.Therefore, follow claims of embodiment and incorporate this embodiment thus clearly into, wherein each claim itself is as independent embodiment of the present invention.
Those skilled in the art are appreciated that and can adaptively change and they are arranged in one or more equipment different with this embodiment the module in the equipment among the embodiment.Can become a module or unit or assembly to the module among the embodiment or unit or combination of components, and can be divided into a plurality of submodules or subelement or sub-component to them in addition.In such feature and/or process or unit at least some are mutually repelling, and can adopt any combination to disclosed all features in this instructions (comprising claim, summary and the accompanying drawing followed) and so all processes or the unit of disclosed any method or equipment make up.Unless clearly statement in addition, disclosed each feature can be by providing identical, being equal to or the alternative features of similar purpose replaces in this instructions (comprising claim, summary and the accompanying drawing followed).
In addition, those skilled in the art can understand, although embodiment more described herein comprise some feature rather than further feature included among other embodiment, the combination of features of different embodiment means and is within the scope of the present invention and forms different embodiment.For example, in the following claims, the one of any of embodiment required for protection can be used with array mode arbitrarily.
Each parts embodiment of the present invention can realize with hardware, perhaps realizes with the software module of moving at one or more processor, and perhaps the combination with them realizes.It will be understood by those of skill in the art that and to use microprocessor or digital signal processor (DSP) to realize according to some or all some or repertoire of parts in the job scheduling device of the embodiment of the invention in practice.The present invention can also be embodied as for part or all equipment or the device program (for example, computer program and computer program) of carrying out method as described herein.Such realization program of the present invention can be stored on the computer-readable medium, perhaps can have the form of one or more signal.Such signal can be downloaded from internet website and obtain, and perhaps provides at carrier signal, perhaps provides with any other form.
It should be noted above-described embodiment the present invention will be described rather than limit the invention, and those skilled in the art can design alternative embodiment under the situation of the scope that does not break away from claims.In the claims, any reference symbol between bracket should be configured to limitations on claims.Word " comprises " not to be got rid of existence and is not listed in element or step in the claim.Being positioned at word " " before the element or " one " does not get rid of and has a plurality of such elements.The present invention can realize by means of the hardware that includes some different elements and by means of the computing machine of suitably programming.In having enumerated the unit claim of some devices, several in these devices can be to come imbody by same hardware branch.Any order is not represented in the use of word first, second and C grade.Can be title with these word explanations.
Claims (10)
1. job scheduling method, wherein, this method comprises:
The resource division of cloud computing platform is become the N group, with the operation submitted to the assignment of allocation when submitting in the group of correspondence; Wherein N is natural number; Each operation utilizes the resource of corresponding group when operation;
For each group,
If the resource of this group can satisfy the demand of the whole operations that are submitted to this group, then directly dispatch the whole operations in this group;
If the resource of this group can not satisfy the demand of the whole operations that are submitted to this group, the operation that then will be submitted to this group is put into the formation of this group and is waited in line, and according to the priority of each operation in this group with wait in line the time and dispatch.
2. the method for claim 1, wherein described priority and the time of waiting in line according to each operation in this group dispatches and comprises:
For each operation in this group arranges the priority weights, priority is more high, and the priority weights are more big;
For each operation in this group arranges the waiting time weights, waiting time is more long, waiting time weights more big;
The priority weights and waiting time weights and be the comprehensive weights of operation;
The comprehensive big operation of weights during priority scheduling should be organized.
3. method as claimed in claim 1 or 2, wherein, this method further comprises:
For each group,
Estimate the operation that can not postponed the working time of the next operation that need dispatch of this group working time, then this operation of priority scheduling if exist during this forms a team to be listed as;
Wherein, operation estimates the working time of portable band when operation is submitted to.
4. method as claimed in claim 3, wherein, this method further comprises:
Estimate working time if surpassed it working time of an operation, then finish the operation of this operation.
5. method as claimed in claim 1 or 2, wherein, this method further comprises:
For periodic real time job, reserve resource requirement at its periodic execution time window for this operation;
If this periodic real time job is not activated in its periodic execution time window, then no longer reserve resource requirement for it.
6. job scheduling device, wherein, this device comprises: grouped element and press the weights scheduling unit;
Described grouped element, the resource division that is suitable for cloud computing platform becomes the N group, with the operation submitted to the assignment of allocation when submitting in the group of correspondence; Wherein N is natural number; Each operation utilizes the resource of corresponding group when operation;
The described weights scheduling unit of pressing is suitable for each group if the resource of this group can satisfy the demand of the whole operations that are submitted to this group, is then directly dispatched the whole operations in this group; If the resource of this group can not satisfy the demand of the whole operations that are submitted to this group, the operation that then will be submitted to this group is put into the formation of this group and is waited in line, and according to the priority of each operation in this group with wait in line the time and dispatch.
7. device as claimed in claim 6, wherein,
The described weights scheduling unit of pressing, be suitable for each group, for each operation in this group arranges the priority weights, priority is more high, and the priority weights are more big, for each operation in this group arranges the waiting time weights, waiting time is more long, waiting time, weights were more big, and with the priority weights and waiting time weights and be the comprehensive weights of operation, and priority scheduling should group in the comprehensive big operation of weights.
8. as claim 6 or 7 described devices, wherein, this device further comprises: by the estimated time scheduling unit;
Described by the estimated time scheduling unit, be suitable for for each group, if this exist in forming a team to be listed as estimate working time can be with should group next operation that needs postponement working time of the operation that dispatch, then this operation of priority scheduling; Wherein, operation estimates the working time of portable band when operation is submitted to.
9. device as claimed in claim 8, wherein,
Described by the estimated time scheduling unit, be suitable for when having surpassed it and estimate working time the working time of an operation, finishing the operation of this operation.
10. as claim 6 or 7 described devices, wherein, this device further comprises: press the periodic scheduling unit;
The described periodic scheduling unit of pressing, be suitable for periodic real time job, reserve resource requirement at its periodic execution time window for this operation, and, if this periodic real time job is not activated in its periodic execution time window, then no longer reserve resource requirement for it.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013102475903A CN103336714A (en) | 2013-06-20 | 2013-06-20 | Operation scheduling method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013102475903A CN103336714A (en) | 2013-06-20 | 2013-06-20 | Operation scheduling method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103336714A true CN103336714A (en) | 2013-10-02 |
Family
ID=49244889
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013102475903A Pending CN103336714A (en) | 2013-06-20 | 2013-06-20 | Operation scheduling method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103336714A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103699448A (en) * | 2014-01-14 | 2014-04-02 | 浪潮(北京)电子信息产业有限公司 | Scheduling method based on time limit and budget in cloud computing environment |
| CN103778011A (en) * | 2014-01-21 | 2014-05-07 | 国家电网公司 | Method for fairly and randomly scheduling tasks of service platform in electric power system |
| CN104217292A (en) * | 2014-09-03 | 2014-12-17 | 南通北城科技创业管理有限公司 | Service platform system based on cloud computing |
| CN104239154A (en) * | 2014-10-09 | 2014-12-24 | 浪潮(北京)电子信息产业有限公司 | Job scheduling method in Hadoop cluster and job scheduler |
| CN104750556A (en) * | 2015-04-14 | 2015-07-01 | 浪潮电子信息产业股份有限公司 | Method and device for dispatching HPC (high performance computing) cluster work |
| CN104778080A (en) * | 2014-01-14 | 2015-07-15 | 中兴通讯股份有限公司 | Job scheduling processing method and device based on coprocessor |
| CN105718316A (en) * | 2014-12-01 | 2016-06-29 | 中国移动通信集团公司 | Job scheduling method and apparatus |
| WO2016197716A1 (en) * | 2016-01-18 | 2016-12-15 | 中兴通讯股份有限公司 | Task scheduling method and device |
| CN107203422A (en) * | 2016-08-28 | 2017-09-26 | 深圳晶泰科技有限公司 | A kind of job scheduling method towards high-performance calculation cloud platform |
| CN107888660A (en) * | 2017-10-13 | 2018-04-06 | 杭州朗和科技有限公司 | Cloud service resource allocation method, medium, device and computing device |
| CN108446169A (en) * | 2017-02-16 | 2018-08-24 | 阿里巴巴集团控股有限公司 | A kind of job scheduling method and device |
| CN109271236A (en) * | 2018-09-18 | 2019-01-25 | 郑州云海信息技术有限公司 | A kind of method, apparatus of traffic scheduling, computer storage medium and terminal |
| CN109947532A (en) * | 2019-03-01 | 2019-06-28 | 中山大学 | A kind of big data method for scheduling task in education cloud platform |
| CN110012507A (en) * | 2019-04-02 | 2019-07-12 | 华南理工大学 | A kind of car networking resource allocation methods that user experience is preferential and system |
| CN110134520A (en) * | 2019-05-27 | 2019-08-16 | 眸芯科技（上海）有限公司 | The application method and system of integrated circuit scarce resource based on queuing |
| CN110673944A (en) * | 2018-07-03 | 2020-01-10 | 杭州海康威视数字技术股份有限公司 | Method and device for executing task |
| CN113760493A (en) * | 2020-11-05 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Job scheduling method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102426542A (en) * | 2011-10-28 | 2012-04-25 | 中国科学院计算技术研究所 | Resource management system for data center and operation calling method |
| CN102750602A (en) * | 2012-04-20 | 2012-10-24 | 广东电网公司信息中心 | Cloud platform isomerism integration resource management system |
| CN102866924A (en) * | 2012-09-12 | 2013-01-09 | 北京航空航天大学 | Method and device for scheduling content integration engine |
| CN102946429A (en) * | 2012-11-07 | 2013-02-27 | 浪潮电子信息产业股份有限公司 | High-efficiency dynamic resource scheduling method based on cloud storage |
| CN103092698A (en) * | 2012-12-24 | 2013-05-08 | 中国科学院深圳先进技术研究院 | System and method of cloud computing application automatic deployment |
-
2013
- 2013-06-20 CN CN2013102475903A patent/CN103336714A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102426542A (en) * | 2011-10-28 | 2012-04-25 | 中国科学院计算技术研究所 | Resource management system for data center and operation calling method |
| CN102750602A (en) * | 2012-04-20 | 2012-10-24 | 广东电网公司信息中心 | Cloud platform isomerism integration resource management system |
| CN102866924A (en) * | 2012-09-12 | 2013-01-09 | 北京航空航天大学 | Method and device for scheduling content integration engine |
| CN102946429A (en) * | 2012-11-07 | 2013-02-27 | 浪潮电子信息产业股份有限公司 | High-efficiency dynamic resource scheduling method based on cloud storage |
| CN103092698A (en) * | 2012-12-24 | 2013-05-08 | 中国科学院深圳先进技术研究院 | System and method of cloud computing application automatic deployment |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103699448A (en) * | 2014-01-14 | 2014-04-02 | 浪潮(北京)电子信息产业有限公司 | Scheduling method based on time limit and budget in cloud computing environment |
| CN104778080A (en) * | 2014-01-14 | 2015-07-15 | 中兴通讯股份有限公司 | Job scheduling processing method and device based on coprocessor |
| CN103778011A (en) * | 2014-01-21 | 2014-05-07 | 国家电网公司 | Method for fairly and randomly scheduling tasks of service platform in electric power system |
| CN103778011B (en) * | 2014-01-21 | 2017-02-08 | 国家电网公司 | Method for fairly and randomly scheduling tasks of service platform in electric power system |
| CN104217292A (en) * | 2014-09-03 | 2014-12-17 | 南通北城科技创业管理有限公司 | Service platform system based on cloud computing |
| CN104239154A (en) * | 2014-10-09 | 2014-12-24 | 浪潮(北京)电子信息产业有限公司 | Job scheduling method in Hadoop cluster and job scheduler |
| CN104239154B (en) * | 2014-10-09 | 2017-08-11 | 浪潮(北京)电子信息产业有限公司 | Job scheduling method and job scheduler in a kind of Hadoop clusters |
| CN105718316A (en) * | 2014-12-01 | 2016-06-29 | 中国移动通信集团公司 | Job scheduling method and apparatus |
| CN104750556A (en) * | 2015-04-14 | 2015-07-01 | 浪潮电子信息产业股份有限公司 | Method and device for dispatching HPC (high performance computing) cluster work |
| WO2016197716A1 (en) * | 2016-01-18 | 2016-12-15 | 中兴通讯股份有限公司 | Task scheduling method and device |
| CN107203422B (en) * | 2016-08-28 | 2020-09-01 | 深圳晶泰科技有限公司 | Job scheduling method for high-performance computing cloud platform |
| CN107203422A (en) * | 2016-08-28 | 2017-09-26 | 深圳晶泰科技有限公司 | A kind of job scheduling method towards high-performance calculation cloud platform |
| CN108446169A (en) * | 2017-02-16 | 2018-08-24 | 阿里巴巴集团控股有限公司 | A kind of job scheduling method and device |
| CN108446169B (en) * | 2017-02-16 | 2022-04-26 | 阿里巴巴集团控股有限公司 | Job scheduling method and device |
| CN107888660A (en) * | 2017-10-13 | 2018-04-06 | 杭州朗和科技有限公司 | Cloud service resource allocation method, medium, device and computing device |
| CN107888660B (en) * | 2017-10-13 | 2021-06-18 | 杭州朗和科技有限公司 | Cloud service resource allocation method, medium, device and computing equipment |
| CN110673944A (en) * | 2018-07-03 | 2020-01-10 | 杭州海康威视数字技术股份有限公司 | Method and device for executing task |
| CN109271236A (en) * | 2018-09-18 | 2019-01-25 | 郑州云海信息技术有限公司 | A kind of method, apparatus of traffic scheduling, computer storage medium and terminal |
| CN109947532A (en) * | 2019-03-01 | 2019-06-28 | 中山大学 | A kind of big data method for scheduling task in education cloud platform |
| CN109947532B (en) * | 2019-03-01 | 2023-06-09 | 中山大学 | Big data task scheduling method in education cloud platform |
| CN110012507A (en) * | 2019-04-02 | 2019-07-12 | 华南理工大学 | A kind of car networking resource allocation methods that user experience is preferential and system |
| CN110012507B (en) * | 2019-04-02 | 2021-01-26 | 华南理工大学 | Internet of vehicles resource allocation method and system with priority of user experience |
| CN110134520A (en) * | 2019-05-27 | 2019-08-16 | 眸芯科技（上海）有限公司 | The application method and system of integrated circuit scarce resource based on queuing |
| CN113760493A (en) * | 2020-11-05 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Job scheduling method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103336714A (en) | Operation scheduling method and device | |
| US9277003B2 (en) | Automated cloud workload management in a map-reduce environment | |
| CN101727357B (en) | Method and apparatus for allocating resources in a compute farm | |
| US9128763B2 (en) | System and method for job scheduling optimization | |
| US9223627B2 (en) | Management of task allocation in a multi-core processing system | |
| CN101499019B (en) | Carrier-grade Ethernet system and real-time task scheduling method used for the same | |
| US20060206894A1 (en) | Method of scheduling jobs using database management system for real-time processing | |
| CN106980532A (en) | A kind of job scheduling method and device | |
| US8539495B2 (en) | Recording medium storing therein a dynamic job scheduling program, job scheduling apparatus, and job scheduling method | |
| CN111338791A (en) | Method, device and equipment for scheduling cluster queue resources and storage medium | |
| Singh | An algorithm to reduce the time complexity of earliest deadline first scheduling algorithm in real-time system | |
| KR101770191B1 (en) | Resource allocation and apparatus | |
| CN103297499A (en) | Scheduling method and system based on cloud platform | |
| CN104199739A (en) | Speculation type Hadoop scheduling method based on load balancing | |
| CN109800092A (en) | A kind of processing method of shared data, device and server | |
| JPH0816410A (en) | Unit and method for schedule control | |
| US8087028B2 (en) | Computing device resource scheduling | |
| US20230289214A1 (en) | Intelligent task messaging queue management | |
| CN109446641A (en) | A kind of multistage Reliability modeling analysis method of cloud computing service system | |
| US11501229B2 (en) | System and method for queue look ahead to optimize work assignment to available agents | |
| CN103593606A (en) | Contextual information managing method and system | |
| Singh et al. | Private cloud scheduling with SJF, bound waiting, priority and load balancing | |
| Postoaca et al. | h-Fair: asymptotic scheduling of heavy workloads in heterogeneous data centers | |
| US11514378B2 (en) | System and method for queue look ahead to optimize agent assignment and utilization | |
| CN117421109B (en) | Training task scheduling method and device, computer 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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20131002 |
|
| RJ01 | Rejection of invention patent application after publication |