CN107291533A - Determine method, the device of upstream node bottleneck degree and system bottleneck degree - Google Patents
Determine method, the device of upstream node bottleneck degree and system bottleneck degree Download PDFInfo
- Publication number
- CN107291533A CN107291533A CN201610197115.3A CN201610197115A CN107291533A CN 107291533 A CN107291533 A CN 107291533A CN 201610197115 A CN201610197115 A CN 201610197115A CN 107291533 A CN107291533 A CN 107291533A
- Authority
- CN
- China
- Prior art keywords
- node
- calculating task
- leaf
- moment
- upstream
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
Abstract
The embodiment of the present application discloses a kind of method, device for determining upstream node bottleneck degree and system bottleneck degree, wherein it is determined that the method for upstream node bottleneck degree, including:Determine all upstream nodes of present node;According to bringing into operation the moment for the moment calculating task corresponding with present node that brings into operation of upstream node correspondence calculating task, it is determined that time range can be optimized;Time range can be optimized, multiple time intervals are split as;Time interval, the time interval run according to each upstream node correspondence calculating task runs the total number of calculating task and can optimize time range, calculates the bottleneck degree of each upstream node;Using the present processes and device, it may be determined that each upstream node starts to perform the restriction degree at moment to present node correspondence calculating task, consequently facilitating the output moment of optimization present node correspondence calculating task.
Description
Technical field
The application is related to the field of internet, more particularly to a kind of method for determining upstream node bottleneck degree and system bottleneck degree,
Device.
Background technology
In scheduling system, scheduling topological diagram (i.e. workflow) is typically provided with, the scheduling system can be according to scheduling topology
Figure, calls application program, performs calculating task.Such as, topological diagram is dispatched, as shown in figure 1, the scheduling system will
Application program is called first, is performed the corresponding calculating task of A nodes, has then been performed in the corresponding calculating task of A nodes
Cheng Hou, recalls application program, performs B node and the corresponding calculating task of C nodes, the like, until performing extremely
Untill D nodes and E nodes.
Because in scheduling system, the output moment (terminating the time of running) of each calculating task is an important indicator,
Therefore, often there is the demand at optimization calculating task output moment;Due to output moment=calculating task of each calculating task
+ the operation duration of calculating task that brings into operation the moment, therefore the moment and can subtract from bringing into operation for calculating task in advance
The aspect of operation duration two of few calculating task, optimizes the output moment of calculating task.
In actual applications, code is write by optimization execution calculating task application program, you can reduce calculating task
Operation duration;And in scheduling system, due to each calculating task bring into operation be constantly in scheduled topological diagram on
That swims node correspondence calculating task terminates what the time of running restricted, such as the beginning of the D nodes correspondence calculating task in Fig. 1
The time of running was restricted by the end time of running of tri- node correspondence calculating tasks of A, B, C of its upstream, that is, was only had
The corresponding calculating task of tri- upstream nodes of A, B, C has been performed both by, and the corresponding calculating task of D nodes can just start to perform.
Further, due to each upstream node be to the restriction degree of current calculating task it is different, therefore, in the prior art,
A kind of method, device for calculating upstream node bottleneck degree and system bottleneck degree is needed badly, to determine each upstream node to current
Calculating task starts to perform the restriction degree at moment, consequently facilitating optimizing the output moment of current calculating task.
Apply for content
A kind of method, device for determining upstream node bottleneck degree and system bottleneck degree is provided in the embodiment of the present application, with
It is determined that each upstream node starts to perform the restriction degree at moment to present node correspondence calculating task, consequently facilitating optimization is worked as
The output moment of front nodal point correspondence calculating task.
In order to solve the above-mentioned technical problem, the embodiment of the present application discloses following technical scheme:
On the one hand, a kind of method for determining upstream node bottleneck degree of disclosure, including:
Determine all upstream nodes of present node;
According to the beginning of the moment that the brings into operation calculating task corresponding with present node of upstream node correspondence calculating task
The time of running, it is determined that time range can be optimized;
Time range can be optimized, multiple time intervals are split as;
Calculating task is run according to time interval, the time interval that each upstream node correspondence calculating task is run
Total number and time range can be optimized, calculate the bottleneck degree of each upstream node.
Optionally, appointed according to calculate corresponding with present node of the moment that brings into operation of upstream node correspondence calculating task
That is engaged in brings into operation the moment, it is determined that time range can be optimized, including:
Judge to whether there is timing calculating task in upstream node correspondence calculating task;
If it does, bringing into operation the timing calculating task moment, brings into operation the moment as target;
If it does not, in upstream node, calculating task bringing into operation the moment earliest is determined, is that target starts fortune
The row moment;
Brought into operation the bringing into operation the moment of moment calculating task corresponding with present node according to target, it is determined that when can optimize
Between scope.
Optionally, methods described also includes:
All upstream nodes of present node are ranked up by bottleneck degree size;
By the upstream node of predetermined number, the upstream bottlenecks node of present node is used as.
Optionally, methods described also includes:
There to be multiple calculating tasks while the time interval run, interval as indirect bottleneck;
According to indirect bottleneck is interval and optimizable time range, the optimization difficulty of present node is determined.
On the other hand, a kind of method for determining system bottleneck degree is also disclosed in the application, including:
Calculate in scheduling topological diagram that a non-leaf nodes is to the shortest path of each leaf node, the non-leaf nodes is
There is the node of downstream node in the scheduling topological diagram;
When non-leaf nodes correspondence calculating task increase operation preset duration, according to shortest path, it is determined that each
The increased duration of calculating task operation institute corresponding to leaf node;
According to the increased duration of each leaf node operation institute, the system bottleneck degree of the non-leaf nodes is determined.
Optionally, calculate scheduling topological diagram in a non-leaf nodes to each leaf node shortest path, including:
In scheduling topological diagram, all paths of a non-leaf nodes a to leaf node are determined;Wherein, each path
Including multiple nodes;
The end time of running of calculating task is corresponded to according to adjacent node in each path and brought into operation the moment, it is determined that often
The path length in individual path;
Path length most short path is determined, is shortest path.
Optionally, when non-leaf nodes correspondence calculating task increase operation preset duration, according to shortest path,
It is determined that the increased duration of calculating task operation institute corresponding to each leaf node, including:
Judge present node correspondence calculating task, whether the increased preset duration of operation institute is more than present node to a leaf
The length of node shortest path;
If it does, determining a length of when the leaf node correspondence calculating task operation increases:Present node is to accrued
Calculate task run increased preset duration-shortest path length;
If less than equal to determining a length of zero when leaf node correspondence calculating task operation institute is increased.
Optionally, methods described also includes:
All non-leaf nodes in topological diagram will be dispatched, be ranked up by system bottleneck degree size;
By the non-leaf nodes of predetermined number, system bottleneck node is used as.
A kind of method for determining system bottleneck degree is also disclosed in another aspect, the application, including:
Determine a non-leaf nodes;
When non-leaf nodes correspondence calculating task operation reduces preset duration, calculate and be activated in scheduling topological diagram
The number of leaf node, the non-leaf nodes is described to be swashed to there is the node of downstream node in the scheduling topological diagram
The leaf node that flap child node changes for the moment that brings into operation of correspondence calculating task with the end time of running;
According to the number and preset duration of the leaf node that is activated, the system bottleneck of the non-leaf nodes is determined
Degree.
Optionally, when non-leaf nodes correspondence calculating task operation reduces preset duration, scheduling topological diagram is calculated
In be activated the number of leaf node, including:
When dispatching non-leaf nodes correspondence calculating task operation reduction preset duration in topological diagram, recalculate described
The end time of running of calculating task;
It is activation node to set the non-leaf nodes;
Judge the end time of running of the activation node correspondence calculating task, if influence is described to activate the direct of node
Downstream node correspondence calculating task brings into operation the moment, the direct downstream node be in the scheduling topological diagram, with
The downstream node that activation node is joined directly together;
If influence, determine the direct downstream node and appoint to activate node, and recalculating activation node correspondence and calculate
That is engaged in brings into operation moment and the end time of running;
Judge whether the activation node is leaf node;
If it is, determining that the activation node is the leaf node that is activated;
If not, returning to circulation performs the end time of running for calculating the activation node correspondence calculating task, if shadow
Ring it is described activation node direct downstream node correspondence calculating task bring into operation the moment the step for.
Optionally, the end time of running of the activation node correspondence calculating task is calculated, if the influence activation section
Moment when bringing into operation of the direct downstream node correspondence calculating task of point, including:
Scheduling topological diagram in, it is determined that activation node direct downstream node, and, the direct downstream node it is straight
Connect upstream node;
In all nodes immediately upstream of a direct downstream node, it is determined that during the end operation of correspondence calculating task the latest
Carve;
Whether the end time of running described in judging the latest is identical with activating the end time of running that node is recalculated;
If identical, it is determined that activation node influences bringing into operation the moment for direct downstream node correspondence calculating task, otherwise,
It is determined that activation node does not influence bringing into operation the moment for direct downstream node correspondence calculating task.
Optionally, methods described also includes:
All non-leaf nodes in topological diagram will be dispatched, be ranked up by system bottleneck degree size;
By the non-leaf nodes of predetermined number, system bottleneck node is used as.
On the other hand, a kind of device for determining upstream node bottleneck degree is also disclosed in the application, including:
Upstream node module is determined, all upstream nodes for determining present node;
It is determined that time range module can be optimized, for bringing into operation the moment according to upstream node correspondence calculating task
Calculating task corresponding with present node brings into operation the moment, it is determined that can optimize time range;
Module is split, for time range can be optimized, multiple time intervals are split as;
Upstream node bottleneck degree computing module, for the time zone run according to each upstream node correspondence calculating task
Between, time interval run the total number of calculating task and time range can be optimized, calculate the bottleneck of each upstream node
Degree.
Optionally, it is determined that time range module can be optimized, including:
Timing calculating task judging unit, appoints for judging to calculate with the presence or absence of timing in upstream node correspondence calculating task
Business;
First object brings into operation moment determining unit, for when there is timing calculating task, the timing to be calculated
Bringing into operation the moment for task, brings into operation the moment as target;
Second target brings into operation moment determining unit, for when in the absence of timing calculating task, in upstream node,
Calculating task bringing into operation the moment earliest is determined, is that target brings into operation the moment;
It is determined that time range unit can be optimized, for being brought into operation moment calculating task corresponding with present node according to target
Bring into operation the moment, it is determined that time range can be optimized.
Optionally, described device also includes:
First order module, for all upstream nodes of present node to be ranked up by bottleneck degree size;
Upstream bottlenecks node determining module, for by the upstream node of predetermined number, being used as the upstream bottlenecks of present node
Node.
Optionally, described device also includes:
Indirect bottleneck interval determination unit, for the time interval that will there are multiple calculating tasks to run simultaneously, as indirect
Bottleneck is interval;
Optimize difficulty determining unit, for according to indirect bottleneck is interval and optimizable time range, it is determined that working as prosthomere
The optimization difficulty of point.
A kind of device for determining system bottleneck degree is also disclosed in another aspect, the application, including:
Shortest path computing module, for the most short of a non-leaf nodes in calculating scheduling topological diagram to each leaf node
Path, the non-leaf nodes dispatches the node that there is downstream node in topological diagram to be described;
Leaf node operation duration determining module, for being preset when non-leaf nodes correspondence calculating task increase operation
During duration, according to shortest path, it is determined that the increased duration of calculating task operation institute corresponding to each leaf node;
The first system bottleneck degree determining module, for according to the increased duration of each leaf node operation institute, it is determined that described
The system bottleneck degree of non-leaf nodes.
Optionally, shortest path computing module, including:
All path determining units, in scheduling topological diagram, determining the institute of a non-leaf nodes a to leaf node
There is path;Wherein, each path includes multiple nodes;
Path length determining unit, for the end time of running according to adjacent node correspondence calculating task in each path
With the moment that brings into operation, it is determined that the path length in each path;
Shortest path determining unit, is shortest path for determining path length most short path.
Optionally, leaf node operation duration determining module, including:
First judging unit, for judging present node correspondence calculating task, whether the increased preset duration of operation institute is big
In the length of present node to a leaf node shortest path;
First operation duration determining unit, for present node correspondence calculating task, the increased preset duration of operation institute is big
When the length of present node to a leaf node shortest path, determine that the leaf node correspondence calculating task operation is increased
Added-time is a length of:Present node correspondence calculating task operation increased preset duration-shortest path length;
Second operation duration determining unit, for present node correspondence calculating task, the increased preset duration of operation institute is small
When length equal to present node to a leaf node shortest path, the leaf node correspondence calculating task operation is determined
A length of zero when institute is increased.
Optionally, described device also includes:
Second order module, for all non-leaf nodes that will be dispatched in topological diagram, is carried out by system bottleneck degree size
Sequence;
The first system bottleneck determining module, for by the non-leaf nodes of predetermined number, being used as system bottleneck node.
On the other hand, a kind of device for determining system bottleneck degree is also disclosed in the application, including:
Non-leaf nodes determining module, for determining a non-leaf nodes;
Calculating is activated leaf node number module, for when non-leaf nodes correspondence calculating task operation reduce it is pre-
If during duration, calculating the number for the leaf node that is activated in scheduling topological diagram, the non-leaf nodes is the scheduling topology
There is the node of downstream node, bring into operation moment and the end that leaf node is correspondence calculating task that be activated in figure
The leaf node that the time of running changes;
Computing system bottleneck degree module, for the number and preset duration of the leaf node that is activated according to, it is determined that
The system bottleneck degree of the non-leaf nodes.
Optionally, the leaf node number module that is activated is calculated, including:
First recalculates unit, for pre- when dispatching non-leaf nodes correspondence calculating task operation reduction in topological diagram
If during duration, recalculating the end time of running of the calculating task;
Setting unit, for setting the non-leaf nodes to be activation node;
Judging unit is influenceed, the end time of running for judging the activation node correspondence calculating task, if influence
The direct downstream node of the activation node corresponds to bringing into operation the moment for calculating task, and the direct downstream node is in institute
State in scheduling topological diagram, the downstream node being joined directly together with activation node;
Node determining unit is activated, if for influenceing, it is activation node to determine the direct downstream node, and again
Calculate bringing into operation moment and the end time of running for activation node correspondence calculating task;
Leaf node determining unit, for judging whether the activation node is leaf node;
Be activated leaf node determining unit, for when activation node is leaf node, determining that the activation node is
Be activated leaf node.
Optionally, judging unit is influenceed, including:
Direct downstream and immediately upstream determination subelement, in scheduling topological diagram, it is determined that activation node it is direct under
Node is swum, and, the node immediately upstream of the direct downstream node;
Terminate time of running determination subelement the latest, in all nodes immediately upstream of a direct downstream node,
It is determined that the end time of running of correspondence calculating task the latest;
Identical judgment sub-unit, for judging the end time of running the latest with activating the end that node is recalculated
Whether the time of running is identical;
Determining unit is influenceed, the end time of running recalculated for the end time of running that ought be the latest and activation node
When identical, it is determined that activation node influences bringing into operation the moment for direct downstream node correspondence calculating task;
Determining unit is not influenceed, for when the end that the end time of running the latest is recalculated with activation node is run
When quarter differs, it is determined that activation node does not influence bringing into operation the moment for direct downstream node correspondence calculating task.
Optionally, described device also includes:
3rd order module, for all non-leaf nodes that will be dispatched in topological diagram, is carried out by system bottleneck degree size
Sequence;
Second system bottleneck determining module, for by the non-leaf nodes of predetermined number, being used as system bottleneck node.
From above technical scheme, in the embodiment of the present application, first in scheduling topological diagram, present node is determined
All upstream nodes;Then according to the moment that the brings into operation calculating corresponding with present node of upstream node correspondence calculating task
Task brings into operation the moment, it is determined that can optimize time range;Time range can subsequently be optimized, when being split as multiple
Between it is interval, determine that the time interval that each upstream node correspondence calculating task is run, and the time interval are run
The total number of calculating task;Finally, time interval, the time zone that calculating task is run are corresponded to according to each upstream node
Between run the total number of calculating task and time range can be optimized, calculate the bottleneck degree of each upstream node.Due to
In the embodiment of the present application, the bottleneck degree of upstream node is bigger, represent its to present node correspondence calculating task start perform when
The restriction degree at quarter is bigger, therefore using method and device disclosed in the present application, can calculate upstream node to present node pair
Calculating task is answered to start to perform the restriction degree at moment, consequently facilitating during the output of optimization present node correspondence calculating task
Carve.
Brief description of the drawings
, below will be to embodiment or existing in order to illustrate more clearly of the embodiment of the present application or technical scheme of the prior art
There is the accompanying drawing used required in technology description to be briefly described, it should be apparent that, for those of ordinary skill in the art
For, without having to pay creative labor, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 is the schematic diagram of the scheduling topological diagram disclosed in the embodiment of the present application;
Fig. 2 is another schematic diagram of the scheduling topological diagram disclosed in the embodiment of the present application;
Fig. 3 is a schematic diagram of the determination upstream node bottleneck degree method disclosed in the embodiment of the present application;
Fig. 4 is another schematic diagram of the determination upstream node bottleneck degree method disclosed in the embodiment of the present application;
Fig. 5 is the another schematic diagram of the determination upstream node bottleneck degree method disclosed in the embodiment of the present application;
Fig. 6 is a schematic diagram of the determination system bottleneck degree method disclosed in the embodiment of the present application;
Fig. 7 is another schematic diagram of the determination system bottleneck degree method disclosed in the embodiment of the present application;
Fig. 8 is the another schematic diagram of the determination system bottleneck degree method disclosed in the embodiment of the present application;
Fig. 9 is the structural representation of the determination upstream node bottleneck degree device disclosed in the embodiment of the present application;
Figure 10 is another schematic diagram of the determination upstream node bottleneck degree method disclosed in the embodiment of the present application.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present application, the technical scheme in the embodiment of the present application is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only some embodiments of the present application, rather than whole embodiments.
Based on the embodiment in the application, the institute that those of ordinary skill in the art are obtained under the premise of creative work is not made
There is other embodiment, belong to the scope of the application protection.
This application discloses a kind of method for determining upstream node bottleneck degree, it can be calculated in scheduling topological diagram using this method
The upstream node bottleneck degree of all non-root node, the non-root node refers in scheduling topological diagram, there is the section of upstream node
Point (C, D, E, F, G, H, I, J node in such as Fig. 2).The application by by taking the inode shown in Fig. 2 as an example,
The process of detailed description the application, and the method disclosed in the present application for calculating upstream node bottleneck degree, as shown in figure 3,
It at least may include following steps:
Step S31:In scheduling topological diagram, all upstream nodes of present node are determined;
In the embodiment of the present application, first in the scheduling topological diagram shown in Fig. 3, determine what present node I was relied on
All upstream nodes, it is possible to find all upstream nodes that node I is relied on include node A, node B, node C, section
Point D, node E, node F and node G.
Step S32:Appointed according to calculate corresponding with present node of the moment that brings into operation of upstream node correspondence calculating task
That is engaged in brings into operation the moment, it is determined that can optimize time range;
In the embodiment of the present application, step S32 implements process, can be as follows:
A:Judge to whether there is timing calculating task in upstream node correspondence calculating task;
In the embodiment of the present application, it is possible to find in node I upstream node A, B, C, D, E, F and G, node
B is timing calculating task, and its moment timing that brings into operation is 5:00 point.
B:If there is timing calculating task, bringing into operation the moment for the timing calculating task starts as target
The time of running;
In the embodiment of the present application, due to no matter how to be optimized to the corresponding calculating task of present node, it starts
The time of running is always later than bringing into operation the moment for its upstream node timing calculating task, therefore need to opening timing calculating task
Time of running beginning brings into operation the moment as target.In the embodiment of the present application, can be by B node correspondence timing calculating task
Bring into operation the moment (5:00 point), brought into operation the moment as target.
C:If there is no timing calculating task, in upstream node, calculating task bringing into operation the moment earliest is determined,
Brought into operation the moment for target;
D:Brought into operation the bringing into operation the moment of moment calculating task corresponding with present node according to target, it is determined that can optimize
Time range.
In the embodiment of the present application, the beginning for the moment calculating task corresponding with present node that specifically target can bring into operation
Time difference between the time of running, as time range can be optimized.In the embodiment of the present application, as shown in Figure 3, save
Bringing into operation for point I is 6 constantly:00, therefore by 6:00 brings into operation the moment 5 with target:Time difference between 00
1 hour (60 minutes), as time range can be optimized.
Step S33:Time range can be optimized, multiple time intervals are split as;
In the embodiment of the present application, user can voluntarily set a time interval, then according to time interval, can optimize
Time range, is split as multiple time intervals.In the embodiment of the present application, can be specifically with 10 minutes for interval, can
Optimization time range (60 minutes) is split as 6 time intervals, respectively 5:00—5:10、5:10—5:20、
5:20—5:30、5:30—5:40、5:40—5:50 and 5:50—6:00.
Step S34:It is determined that the time interval that each upstream node correspondence calculating task is run, and the time interval
The total number of run calculating task;
In the embodiment of the present application, as shown in Figure 3, when upstream A nodes correspondence calculating task does not operate in above-mentioned
Between in area, i.e., the time interval that A nodes correspondence calculating task is run is 0;Upstream B node correspondence calculating task is transported
Capable time interval is 5:00—5:10 and 5:10—5:20;Upstream C nodes correspondence calculating task is run
Time interval 5:00—5:10 and 5:10—5:20;The time zone that upstream D nodes correspondence calculating task is run
Between be 5:00—5:10、5:10—5:20 and 5:20—5:30;Upstream E nodes correspondence calculating task is transported
Capable time interval is 5:20—5:30 and 5:30—5:40;Upstream F nodes correspondence calculating task is run
Time interval is 5:20—5:30、5:30—5:40、5:40—5:50 and 5:50—6:00;Upstream G
The time interval that node is run is 5:30—5:40 and 5:40—5:50.Pass through above-mentioned analysis, 5:
00—5:The calculating task that 10 this time interval are run is the corresponding calculating task of the node of B, C, D tri-, totally 3;
5:10—5:The calculating task that 20 this time interval are run is the corresponding calculating task of the node of B, C, D tri-, totally 3
It is individual;5:20—5:The calculating task that 30 this time interval are run is the corresponding calculating task of the node of D, E, F tri-,
Totally 3;5:30—5:The calculating task that 40 this time interval are run, which is that the node of E, F, G tri- is corresponding, to be calculated
Task, totally 3;5:40—5:The calculating task that 50 this time interval are run, which is that F, G node are corresponding, to be calculated
Task, totally 2;5:50—6:The calculating task that 00 this time interval is run is that corresponding calculate of F nodes is appointed
Business, only 1.
Step S35:Meter is run according to time interval, the time interval that each upstream node correspondence calculating task is run
The total number of calculation task and time range can be optimized, calculate the bottleneck degree of each upstream node.
In the embodiment of the present application, it is assumed that the time interval that upstream node correspondence calculating task is run is time interval
1st, time interval 2 ... time interval i ... time intervals n, then can specifically utilize following upstream node bottleneck degree meters
Formula is calculated, the bottleneck degree of each upstream node is calculated, the upstream bottlenecks degree calculation formula is:
The i and n are just
Integer, and i is less than or equal to n.
In the embodiment of the present application, the example above is still continued to use, passes through above-mentioned discussion, A nodes correspondence calculating task institute
The time interval of operation is 0, therefore upstream node A bottleneck degree is 0.And what B node correspondence calculating task was run
Time interval is 5:00—5:10 and 5:10—5:20, and 5:00—5:10 this time interval are run
The total number of calculating task is 3,5:10—5:The total number that 20 this time interval run calculating task is also 3
It is individual, therefore upstream node B bottleneck degree=[(5:00-5:10)/3+(5:10-5:20)/3]/[5:00-6:00]=[10 points
Clock/3+10 minutes/3]/60 minute=11.1%;Similarly, using the above method, the bottleneck degree that can try to achieve upstream node C is
11.1%, upstream node D bottleneck degree are 16.7%, and upstream node E bottleneck degree is 11.1%, upstream node F bottle
Neck degree is 36.1%, and upstream node G bottleneck degree is 13.9%.
From above technical scheme, in the embodiment of the present application, first in scheduling topological diagram, present node is determined
All upstream nodes;Then according to the moment that the brings into operation calculating corresponding with present node of upstream node correspondence calculating task
Task brings into operation the moment, it is determined that can optimize time range;Time range can subsequently be optimized, when being split as multiple
Between it is interval, determine that the time interval that each upstream node correspondence calculating task is run, and the time interval are run
The total number of calculating task;Finally, time interval, the time zone that calculating task is run are corresponded to according to each upstream node
Between run the total number of calculating task and time range can be optimized, calculate the bottleneck degree of each upstream node.Due to
In the embodiment of the present application, the bottleneck degree of upstream node is bigger, represent its to present node correspondence calculating task start perform when
The restriction degree at quarter is bigger, therefore using method disclosed in the present application, can calculate upstream node and present node correspondence is calculated
Task starts to perform the restriction degree at moment, consequently facilitating the output moment of optimization present node correspondence calculating task.
In another possible embodiments of the application, as shown in figure 4, the method in above-mentioned all embodiments, in addition to:
Step S41:All upstream nodes of present node are ranked up by bottleneck degree size;
In the embodiment of the present application, all upstream nodes of present node can from large to small be sorted by bottleneck degree, also may be used
Sorted from small to large by bottleneck degree.In the embodiment of the present application, the example above is still continued to use, can be by all of present node I
Upstream node A, B, C, D, E, F and G sort from large to small by bottleneck degree, respectively upstream node F, D, G,
B、C、E、A。
Step S42:By the upstream node of predetermined number, the upstream bottlenecks node of present node is used as.
In the embodiment of the present application, a default upstream node that can specifically by bottleneck degree greatly, is used as the upstream of present node
Bottleneck node.The example above is continued to use, bottleneck degree larger node F and node D can be regard as present node I upstream
Bottleneck node.
Therefore, in the embodiment of the present application, it may be determined that the upstream bottlenecks node of present node.
In another possible embodiments of the application, as shown in figure 5, the method in above-mentioned all embodiments, can also be wrapped
Include:
Step S51:There to be multiple calculating tasks while the time interval run, interval as indirect bottleneck;
In the embodiment of the present application, the example above is still continued to use, it may be determined that there are multiple calculating tasks while the time zone run
Between, it is 5:00—5:10、5:10—5:20、5:20—5:30、5:30—5:40、5:40—5:50
And 5:50—6:00;
Step S52:According to indirect bottleneck is interval and optimizable time range, the optimization difficulty of present node is calculated.
In the embodiment of the present application, due to only having the time interval of a upstream node operation, optimization gets up to be easier,
Only the upstream node is optimized, you can optimize bringing into operation the moment for present node;But there are multiple upstream nodes same
Shi Yunhang time interval, i.e., indirect bottleneck is interval, to optimize multiple upstream nodes simultaneously, could optimize present node
Bring into operation the moment, therefore, the optimization difficulty of present node can be characterized with indirect bottleneck interval.
In the embodiment of the present application, when indirect bottleneck interval is indirect bottleneck interval 1, indirect bottleneck interval 2 ... is indirectly
Bottleneck interval p ... indirectly bottleneck interval m when, can specifically using optimization difficulty calculation formula, calculating present node it is excellent
Change difficulty, the optimization difficulty calculation formula is specially:
The p and m are positive integer, and the p is less than or equal to m;
In the embodiment of the present application, using above-mentioned formula, present node I optimization difficulty=[(5 are calculated:00—5:
10)+(5:10—5:20)+(5:20—5:30)+(5:30—5:40)+(5:40—5:50)+(5:50—6:
00)]/[[5:00-6:00]]=[+10 minutes+10 minutes+10 minutes+10 minutes 10 minutes]/60 minutes=83.3%.
Therefore, in the embodiment of the present application, the optimization difficulty of present node can be calculated.
Disclosed herein as well is a kind of method for determining system bottleneck degree, it can calculate scheduling using this method and flutter in figure and own
The system bottleneck degree of non-leaf nodes, the non-leaf nodes refers in scheduling topological diagram, the node (ratio that there is downstream node
Such as A, B, C, D, E, F, G node in Fig. 2).The application by with the non-leaf nodes B shown in Fig. 2 to accrued
Exemplified by calculation task increase operation 1 minute, the process of detailed full and clear the application;And computing system bottleneck disclosed in the present application
The method of degree, as shown in fig. 6, at least comprising the following steps:
Step S61:Calculate scheduling topological diagram in a non-leaf nodes to each leaf node shortest path;
In the embodiment of the present application, to calculate in Fig. 3 exemplified by non-leaf nodes B to leaf node I shortest path,
Describe the step S61 process that implements in detail, and step S61 implements process, can be as follows:
A:In scheduling topological diagram, all paths of a non-leaf nodes a to leaf node are determined;Wherein, Mei Gelu
Footpath includes multiple nodes;
In the embodiment of the present application, it is first determined node B to node I all paths, opened up in the scheduling shown in Fig. 3
Flutter in figure, it is possible to find node B to node I path, including:B-G-I paths, B-I paths and B-E-I roads
Footpath;
B:The end time of running of calculating task is corresponded to according to each path adjacent node and brought into operation the moment, it is determined that often
The path length in individual path;
In the embodiment of the present application, can be specifically by the end time of running of adjacent node correspondence calculating task in each path
Time difference between the moment that brings into operation, it is used as subpath length;Then by all subpath length phases in current path
Plus, it is used as the path length of current path.
In the embodiment of the present application, exemplified by calculating the path length in this path of B-G-H, said process is described in detail:
Calculate node B end time of running (5 first:20) (5 are brought into operation the moment with node G:30) between when
Between poor (be computed can obtain time difference for 10 minutes), and regard the time difference 10 as a subpath length;Then section is calculated
The point G end time of running (5:50) (6 are brought into operation the moment with node I:00) time difference between is (through meter
Calculate and understand that the time difference is also 10 minutes), and it regard the time difference 10 as another subpath length;Finally by current path
All subpath length are added (10+10=20), are used as the path length in this path of B-G-H;Similarly, it can calculate
The path length in this path of B-I is that the path length in 40, B-this path of E-I is 20.
C:Path length most short path is determined, is shortest path.
Because in the above-described embodiments, the path length in B-G-I and the two paths of B-E-I is 20, therefore can be true
It is B-G-I or B-E-I to determine current node B to node I shortest path.
Step S62:When the corresponding calculating task increase operation preset duration of the non-leaf nodes, according to described most short
Path, it is determined that the corresponding increased duration of calculating task operation institute of each leaf node;
In the embodiment of the present application, run 1 minute in being increased with B node correspondence calculating task, calculate leaf node I correspondences
Exemplified by the increased duration of calculating task operation institute, illustrate step S62;In the embodiment of the present application, step S62's is specific
Implementation process is as follows:
A:Judge present node correspondence calculating task, whether the increased preset duration of operation institute is more than present node to a leaf
The shortest path length of child node;
B:If it does, determining a length of when the leaf node correspondence calculating task operation institute is increased:Present node pair
Increased preset duration-the shortest path length of task run institute should be calculated;
C:If less than equal to determining a length of zero when leaf node correspondence calculating task operation institute is increased.
In the embodiment of the present application, the example above is still continued to use, by B node correspondence calculating task operation increases by 1 point
Clock, less than or equal to the shortest path length 20 of B node to inode, performs C;Therefore it can determine that inode correspondence calculating
A length of zero when task run institute is increased;Similarly, can calculate H nodes correspondence calculating task operation institute it is increased when a length of 0
Minute, a length of 1 minute when J node correspondence calculating task operation institute is increased;
In the embodiment of the present application, Route Routes length can represent present node to the buffer time of leaf node, therefore,
When the increased preset duration of present node correspondence calculating task operation institute is less than above-mentioned buffer time, it may be determined that present node
On leaf node and in the absence of any influence.Because in actual applications, calculating task corresponding to leaf node is generally tool
Body application, its output moment is even more important, because in the embodiment of the present application, with each non-leaf nodes to leaf node
Influence be yardstick, weigh the influence of each non-leaf nodes to whole scheduling system.
Step S63:According to the increased duration of each leaf node operation institute, the system bottleneck of the non-leaf nodes is determined
Degree.
In the embodiment of the present application, the operation of each leaf node specifically can be increased into duration to be added, is used as the non-leaf nodes
System bottleneck degree.
In the embodiment of the present application, H in topological diagram, I, the increased duration of J node operation institute can will be dispatched shown in Fig. 3
It is added (0+0+1=1), is used as the system bottleneck degree of inode.
Therefore, using the above method, the system bottleneck degree of each non-leaf nodes in scheduling topological diagram can be calculated.
In another possible embodiments of the application, the method in above-mentioned all embodiments may also include:
All non-leaf nodes in topological diagram will be dispatched, be ranked up by system bottleneck degree size;
By the non-leaf nodes of predetermined number, system bottleneck node is used as.
In the embodiment of the present application, a default non-leaf nodes that can specifically by system bottleneck degree greatly, is used as system bottleneck
Node.
Using the method in above-described embodiment, can calculate in scheduling scheduling topological diagram, all non-leaf nodes (A, B, C,
D, E, F, G node) system bottleneck degree, and in the embodiment of the present application, can be to all non-sons in scheduling topological diagram
The system bottleneck degree of node is ranked up, and by the larger non-leaf nodes of bottleneck degree, is used as system bottleneck node.
The application discloses a kind of method for determining system bottleneck degree again, and scheduling can be equally calculated using this method and is flutterred in figure
The system bottleneck degree of all non-leaf nodes.The application will be reduced with the non-leaf nodes B correspondence calculating tasks shown in Fig. 2
Exemplified by operation 1 minute, the process of detailed full and clear the application;And the method for computing system bottleneck degree disclosed in the present application,
As shown in fig. 7, at least comprising the following steps:
Step S71:In scheduling topological diagram, a non-leaf nodes is determined;
Step S72:When non-leaf nodes correspondence calculating task reduces operation preset duration, scheduling topological diagram is calculated
In be activated the number of leaf node, the non-leaf nodes to there is the node of downstream node in the scheduling topological diagram,
The leaf that the leaf node that is activated changes for the moment that brings into operation of correspondence calculating task with the end time of running
Node;
In the embodiment of the present application, step S72 implements process, can be as shown in Figure 8:
Step S71-1:When dispatching non-leaf nodes correspondence calculating task reduction operation preset duration in topological diagram, weight
Newly calculate the end time of running of the calculating task;
In the embodiment of the present application, first in the scheduling topological diagram shown in Fig. 3, B node correspondence calculating task is determined
Start perform the moment and terminate perform the moment, respectively 5:00 and 5:20;And when B node correspondence calculating task subtracts
When running 1 minute less, the end time of running for determining B node correspondence calculating task is 5:19.
Step S71-2:It is activation node to set the non-leaf nodes;
In the embodiment of the present application, B node is set to activate node;
Step S71-3:Judge the end time of running of the activation node correspondence calculating task, if influence is described to be calculated
The direct downstream node of node corresponds to bringing into operation the moment for calculating task, and the direct downstream node is in scheduling topological diagram
In, the downstream node being joined directly together with activation node;If influence, step S71-4 is performed;If do not influenceed, perform
Step S71-7;
In the embodiment of the present application, step S71-3 implements process, can be as follows:
A:In debugging topological diagram, it is determined that the direct downstream node of activation node, and the direct downstream node are straight
Connect upstream node;
In the embodiment of the present application, first in the scheduling topological diagram shown in Fig. 3, it is determined that activation node B it is direct under
Node is swum, is node G, I and E;Then again in scheduling topological diagram, it is determined that directly downstream node G is immediately upstream
Node, is node B and D, direct downstream node I node immediately upstream is node G, B and E, direct downstream
Node E node immediately upstream, is node B.
B:In all nodes immediately upstream of a direct downstream node, it is determined that the end operation of correspondence calculating task the latest
Moment;
In the embodiment of the present application, by taking direct downstream node G as an example, the process of the application is described in detail, first in Fig. 3
In shown scheduling topological diagram, determine that all upstream nodes (the node B and D) of G nodes brings into operation the moment, can send out
Existing, node B bringing into operation for calculating task of correspondence has turned into 5 constantly:19, and the beginning of node D correspondence calculating tasks
The time of running is 5:30;It can determine that calculating task the latest brings into operation constantly as 5:30.
C:Judge whether the end time of running the latest is identical with activating the end time of running that node is recalculated;If
It is identical, it is determined that performing D, otherwise, perform E;
In the embodiment of the present application, it is possible to find the end time of running 5 the latest:30 knots recalculated with activation node B
The beam time of running 5:19 is not consistent, therefore can determine that node B not to direct downstream node D correspondence calculating tasks
Bring into operation and produce influence constantly.Similarly, using the above method, it may be determined that node B direct downstream node E is impacted,
Inode is simultaneously unaffected.Due in the embodiment of the present application, only appointing when all upstream nodes correspondence of a node is calculated
When business terminates operation, present node correspondence calculating task can just bring into operation, therefore, when node correspondence calculating task
End time of running for recalculating of the time of running the latest and its upstream activat node it is inconsistent when, it may be determined that activation node
The duration of operation is reduced, influence is not produced on present node.
D:It is determined that activation node influences bringing into operation the moment for direct downstream node correspondence calculating task, step is performed
S71-4;
E:It is determined that activation node does not influence bringing into operation the moment for direct downstream node correspondence calculating task, step is performed
S71-7。
Step S71-4:It is determined that directly downstream node is activation node, and recalculate activation node correspondence calculating task
Bring into operation moment and the end time of running;
In the embodiment of the present application, the example above is still continued to use, it may be determined that E nodes are activation node.By shown in Fig. 3
Scheduling topological diagram understand, be changed into 5 when the end time of running of B node:When 19, bringing into operation for E nodes becomes constantly
For 5:19, the end time of running is changed into 5:39.
Step S71-5:Judge whether activation node is leaf node;If it is, performing step S71-6;If not,
Return and perform step S71-3;
In the embodiment of the present application, because E nodes are not leaf node, return and perform step S71-3, using above-mentioned
Method, it may be determined that J node is activation node, and now J node is also leaf node, performs step S71-6, now
J tubercles be can determine that for the section leaf node that is activated.
Step S71-6:It is determined that activation node is the leaf node that is activated, step S72 is performed;
Step S71-7:Determine that activation leaf node is not present in the scheduling topological diagram, perform step S72.
Step S73:According to the number and preset duration of the leaf node that is activated, the non-leaf nodes is determined
System bottleneck degree.
In the embodiment of the present application, can be with specific reference to system bottleneck degree calculation formula, calculate the non-leaf nodes is
Unite bottleneck degree, the system bottleneck degree calculation formula is:System bottleneck degree=be activated leaf node number * preset durations.
In the embodiment of the present application, the example above is still continued to use, it is possible to find, when B node correspondence calculating task reduces operation 1
During minute, in the scheduling topological diagram shown in Fig. 3, the number for the leaf node that is activated is 1 (only J node), because
System bottleneck degree=1*1=1 of this B node.
Therefore, using the above method, it may be determined that in scheduling topological diagram, the system bottleneck degree of each non-leaf nodes.
In another possible embodiments of the application, the method in above-mentioned all embodiments may also include:
All non-leaf nodes in topological diagram will be dispatched, be ranked up by system bottleneck degree size;
By the non-leaf nodes of predetermined number, system bottleneck node is used as.
In the embodiment of the present application, a default non-leaf nodes that can specifically by system bottleneck degree greatly, is used as system bottleneck node.
Using the method in above-described embodiment, can calculate in scheduling scheduling topological diagram, all non-leaf nodes (A, B, C,
D, E, F, G node) system bottleneck degree, and in the embodiment of the present application, can be to all non-sons in scheduling topological diagram
The system bottleneck degree of node is ranked up, and by the larger non-leaf nodes of bottleneck degree, is used as system bottleneck node.
Disclosed herein as well is a kind of method for determining upstream node bottleneck degree, as shown in Figure 10, methods described is at least wrapped
Include following steps:
Step S101:Determine all upstream nodes of present node;
Step S102:According to the moment that the brings into operation calculating corresponding with present node of upstream node correspondence calculating task
Task brings into operation the moment, it is determined that can optimize time range;
Step S103:Time range can be optimized, multiple time intervals are split as;
Step S104:Run according to time interval, the time interval that each upstream node correspondence calculating task is run
The total number of calculating task and time range can be optimized, calculate the bottleneck degree of each upstream node.
From above technical scheme, in the embodiment of the present application, first in scheduling topological diagram, present node is determined
All upstream nodes;Then according to the moment that the brings into operation calculating corresponding with present node of upstream node correspondence calculating task
Task brings into operation the moment, it is determined that can optimize time range;Time range can subsequently be optimized, when being split as multiple
Between it is interval, determine that the time interval that each upstream node correspondence calculating task is run, and the time interval are run
The total number of calculating task;Finally, time interval, the time zone that calculating task is run are corresponded to according to each upstream node
Between run the total number of calculating task and time range can be optimized, calculate the bottleneck degree of each upstream node.Due to
In the embodiment of the present application, the bottleneck degree of upstream node is bigger, represent its to present node correspondence calculating task start perform when
The restriction degree at quarter is bigger, therefore using method disclosed in the present application, can calculate upstream node and present node correspondence is calculated
Task starts to perform the restriction degree at moment, consequently facilitating the output moment of optimization present node correspondence calculating task.
The description of embodiment of the method more than, it is apparent to those skilled in the art that the application can
Realized by the mode of software plus required general hardware platform, naturally it is also possible to by hardware, but in many cases before
Person is more preferably embodiment.Understood based on such, the technical scheme of the application is substantially in other words to prior art
The part contributed can be embodied in the form of software product, and the computer software product is stored in storage Jie
In matter, including some instructions are to cause a computer equipment (can be personal computer, server, or network
Equipment etc.) perform all or part of step of the application each embodiment methods described.And foregoing storage medium includes:
Read-only storage (ROM), random access memory (RAM), magnetic disc or CD etc. are various can be with store program codes
Medium.
Embodiment of the method with the determination upstream node bottleneck degree that the application is provided is corresponding, and present invention also provides one kind
The device of upstream node bottleneck degree is determined, as shown in figure 9, at least including:
Upstream node module 91 is determined, all upstream nodes for determining present node;
It is determined that time range module 92 can be optimized, for according to upstream node when bringing into operation of calculating task of correspondence
Bringing into operation the moment for calculating task corresponding with present node is carved, it is determined that time range can be optimized;
Module 93 is split, for time range can be optimized, multiple time intervals are split as;
Upstream node bottleneck degree computing module 94, for the time run according to each upstream node correspondence calculating task
Interval, time interval runs the total number of calculating task and can optimize time range, calculates the bottle of each upstream node
Neck degree.
Therefore, in the embodiment of the present application, it may be determined that the bottleneck degree of each upstream node.
In another possible embodiments of the application, the determination in above-mentioned all embodiments can optimize time range module,
Including:
Timing calculating task judging unit, appoints for judging to calculate with the presence or absence of timing in upstream node correspondence calculating task
Business;
First object brings into operation moment determining unit, for when there is timing calculating task, the timing to be calculated
Bringing into operation the moment for task, brings into operation the moment as target;
Second target brings into operation moment determining unit, for when in the absence of timing calculating task, in upstream node,
Calculating task bringing into operation the moment earliest is determined, is that target brings into operation the moment;
It is determined that time range unit can be optimized, for being brought into operation moment calculating task corresponding with present node according to target
Bring into operation the moment, it is determined that time range can be optimized.
In another possible embodiments of the application, the device in above-mentioned all embodiments also includes:
First order module, for all upstream nodes of present node to be ranked up by bottleneck degree size;
Upstream bottlenecks node determining module, for by the upstream node of predetermined number, being used as the upstream bottlenecks of present node
Node.
In another possible embodiments of the application, the device in above-mentioned all embodiments also includes:
Indirect bottleneck interval determination unit, for the time interval that will there are multiple calculating tasks to run simultaneously, as indirect
Bottleneck is interval;
Optimize difficulty determining unit, for according to indirect bottleneck is interval and optimizable time range, it is determined that working as prosthomere
The optimization difficulty of point.
With corresponding to the embodiment of the method for determination system bottleneck degree disclosed above, the application is also disclosed a kind of determines system
The device of system bottleneck degree, including:
Shortest path computing module, for the most short of a non-leaf nodes in calculating scheduling topological diagram to each leaf node
Path, the non-leaf nodes dispatches the node that there is downstream node in topological diagram to be described;
Leaf node operation duration determining module, for being preset when non-leaf nodes correspondence calculating task increase operation
During duration, according to shortest path, it is determined that the increased duration of calculating task operation institute corresponding to each leaf node;
The first system bottleneck degree determining module, for according to the increased duration of each leaf node operation institute, it is determined that described
The system bottleneck degree of non-leaf nodes.
Therefore, using said apparatus, the system bottleneck degree of each non-leaf nodes in scheduling topological diagram can be calculated.
In another possible embodiments of the application, the shortest path computing module in above-mentioned all embodiments, including:
All path determining units, in scheduling topological diagram, determining the institute of a non-leaf nodes a to leaf node
There is path;Wherein, each path includes multiple nodes;
Path length determining unit, for the end time of running according to adjacent node correspondence calculating task in each path
With the moment that brings into operation, it is determined that the path length in each path;
Shortest path determining unit, is shortest path for determining path length most short path.、、
In the another possible embodiments of the application, the leaf node operation duration determining module in above-mentioned all embodiments,
Including:
First judging unit, for judging present node correspondence calculating task, whether the increased preset duration of operation institute is big
In the length of present node to a leaf node shortest path;
First operation duration determining unit, for present node correspondence calculating task, the increased preset duration of operation institute is big
When the length of present node to a leaf node shortest path, determine that the leaf node correspondence calculating task operation is increased
Added-time is a length of:Present node correspondence calculating task operation increased preset duration-shortest path length;
Second operation duration determining unit, for present node correspondence calculating task, the increased preset duration of operation institute is small
When length equal to present node to a leaf node shortest path, the leaf node correspondence calculating task operation is determined
A length of zero when institute is increased.
In the another possible embodiments of the application, the device in above-mentioned all embodiments also includes:
Second order module, for all non-leaf nodes that will be dispatched in topological diagram, is carried out by system bottleneck degree size
Sequence;
The first system bottleneck determining module, for by the non-leaf nodes of predetermined number, being used as system bottleneck node.
With corresponding to the embodiment of the method for computing system bottleneck degree disclosed above, the application is also disclosed a kind of calculates system
The device of system bottleneck degree, including:
Non-leaf nodes determining module, for determining a non-leaf nodes;
Calculating is activated leaf node number module, for when non-leaf nodes correspondence calculating task operation reduce it is pre-
If during duration, calculating the number for the leaf node that is activated in the scheduling topological diagram, the non-leaf nodes is the scheduling
There is the node of downstream node in topological diagram, the leaf node that is activated be correspondence calculating task bring into operation the moment and
The leaf node that the end time of running changes;
Computing system bottleneck degree module, for the number and preset duration of the leaf node that is activated according to, it is determined that
The system bottleneck degree of the non-leaf nodes.Therefore, using said apparatus, it can calculate each non-in scheduling topological diagram
The system bottleneck degree of leaf node.
In the another possible embodiments of the application, the calculating in above-mentioned all embodiments is activated leaf node number mould
Block, including:
First recalculates unit, for pre- when dispatching non-leaf nodes correspondence calculating task operation reduction in topological diagram
If during duration, recalculating the end time of running of the calculating task;
Setting unit, for setting the non-leaf nodes to be activation node;
Judging unit is influenceed, the end time of running for judging the activation node correspondence calculating task, if influence
The direct downstream node of the activation node corresponds to bringing into operation the moment for calculating task, and the direct downstream node is in institute
State in scheduling topological diagram, the downstream node being joined directly together with activation node;
Node determining unit is activated, if for influenceing, it is activation node to determine the direct downstream node, and again
Calculate bringing into operation moment and the end time of running for activation node correspondence calculating task;
Leaf node determining unit, for judging whether the activation node is leaf node;
Be activated leaf node determining unit, for when activation node is leaf node, determining that the activation node is
Be activated leaf node.
In the another possible embodiments of the application, the influence judging unit in above-mentioned all embodiments, including:
Direct downstream and immediately upstream determination subelement, in scheduling topological diagram, it is determined that activation node it is direct under
Node is swum, and, the node immediately upstream of the direct downstream node;
Terminate time of running determination subelement the latest, in all nodes immediately upstream of a direct downstream node,
It is determined that the end time of running of correspondence calculating task the latest;
Identical judgment sub-unit, for judging the end time of running the latest with activating the end that node is recalculated
Whether the time of running is identical;
Determining unit is influenceed, the end time of running recalculated for the end time of running that ought be the latest and activation node
When identical, it is determined that activation node influences bringing into operation the moment for direct downstream node correspondence calculating task;
Determining unit is not influenceed, for when the end that the end time of running the latest is recalculated with activation node is run
When quarter differs, it is determined that activation node does not influence bringing into operation the moment for direct downstream node correspondence calculating task.
In another possible embodiments of the application, the device in above-mentioned all embodiments also includes:
3rd order module, for all non-leaf nodes that will be dispatched in topological diagram, is carried out by system bottleneck degree size
Sequence;
Second system bottleneck determining module, for by the non-leaf nodes of predetermined number, being used as system bottleneck node.
It should be noted that herein, the relational terms of such as " first " and " second " or the like are used merely to
By an entity or operation with another entity or operate make a distinction, and not necessarily require or imply these entities or
There is any this actual relation or order between operation.Moreover, term " comprising ", "comprising" or its
What his variant is intended to including for nonexcludability so that process, method, article including a series of key elements or
Person's equipment not only includes those key elements, but also other key elements including being not expressly set out, or also includes to be this
Process, method, article or the intrinsic key element of equipment.In the absence of more restrictions, by sentence " including one
It is individual ... " limit key element, it is not excluded that also exist in the process including the key element, method, article or equipment
Other identical element.
Described above is only the embodiment of the application, is made skilled artisans appreciate that or realizing this Shen
Please.A variety of modifications to these embodiments will be apparent to one skilled in the art, determine herein
The General Principle of justice can in other embodiments be realized in the case where not departing from spirit herein or scope.Cause
This, the application is not intended to be limited to the embodiments shown herein, and is to fit to and principles disclosed herein
The most wide scope consistent with features of novelty.
Claims (24)
1. a kind of method for determining upstream node bottleneck degree, it is characterised in that including:
Determine all upstream nodes of present node;
Fortune is started according to the moment that the brings into operation calculating task corresponding with present node of upstream node correspondence calculating task
The row moment, it is determined that time range can be optimized;
Time range can be optimized, multiple time intervals are split as;
The total of calculating task is run according to time interval, the time interval that each upstream node correspondence calculating task is run
Number and time range can be optimized, calculate the bottleneck degree of each upstream node.
2. according to the method described in claim 1, it is characterised in that according to opening for upstream node correspondence calculating task
The time of running calculating task corresponding with present node that begins brings into operation the moment, it is determined that can optimize time range, including:
Judge to whether there is timing calculating task in upstream node correspondence calculating task;
If it does, bringing into operation the timing calculating task moment, brings into operation the moment as target;
If it does not, in upstream node, calculating task bringing into operation the moment earliest is determined, is that target brings into operation
Moment;
Brought into operation the bringing into operation the moment of moment calculating task corresponding with present node according to target, it is determined that the time can be optimized
Scope.
3. method according to claim 1 or 2, it is characterised in that methods described also includes:
All upstream nodes of present node are ranked up by bottleneck degree size;
By the upstream node of predetermined number, the upstream bottlenecks node of present node is used as.
4. method according to claim 1 or 2, it is characterised in that methods described also includes:
There to be multiple calculating tasks while the time interval run, interval as indirect bottleneck;
According to indirect bottleneck is interval and optimizable time range, the optimization difficulty of present node is determined.
5. a kind of method for determining system bottleneck degree, it is characterised in that including:
A non-leaf nodes is to the shortest path of each leaf node in calculating scheduling topological diagram, and the non-leaf nodes is institute
State the node that there is downstream node in scheduling topological diagram;
When non-leaf nodes correspondence calculating task increase operation preset duration, according to shortest path, it is determined that each leaf
The increased duration of calculating task operation institute corresponding to child node;
According to the increased duration of each leaf node operation institute, the system bottleneck degree of the non-leaf nodes is determined.
6. method according to claim 5, it is characterised in that calculate in scheduling topological diagram a non-leaf nodes to every
The shortest path of individual leaf node, including:
In scheduling topological diagram, all paths of a non-leaf nodes a to leaf node are determined;Wherein, each path bag
Include multiple nodes;
The end time of running of calculating task is corresponded to according to adjacent node in each path and brought into operation the moment, it is determined that each
The path length in path;
Path length most short path is determined, is shortest path.
7. method according to claim 5, it is characterised in that when non-leaf nodes correspondence calculating task increase
When running preset duration, according to shortest path, it is determined that the increased duration of calculating task operation institute corresponding to each leaf node,
Including:
Judge present node correspondence calculating task, whether the increased preset duration of operation institute is more than present node to a leaf section
The length of point shortest path;
If it does, determining a length of when the leaf node correspondence calculating task operation increases:Present node correspondence is calculated
Task run increased preset duration-shortest path length;
If less than equal to determining a length of zero when leaf node correspondence calculating task operation institute is increased.
8. the method according to any one of claim 5-7, it is characterised in that methods described also includes:
All non-leaf nodes in topological diagram will be dispatched, be ranked up by system bottleneck degree size;
By the non-leaf nodes of predetermined number, system bottleneck node is used as.
9. a kind of method for determining system bottleneck degree, it is characterised in that including:
Determine a non-leaf nodes;
When non-leaf nodes correspondence calculating task operation reduces preset duration, the leaf that is activated in scheduling topological diagram is calculated
The number of child node, the non-leaf nodes is described to be activated to there is the node of downstream node in the scheduling topological diagram
The leaf node that leaf node changes for the moment that brings into operation of correspondence calculating task with the end time of running;
According to the number and preset duration of the leaf node that is activated, the system bottleneck degree of the non-leaf nodes is determined.
10. method according to claim 9, it is characterised in that when non-leaf nodes correspondence calculating task fortune
When row reduces preset duration, the number for the leaf node that is activated in scheduling topological diagram is calculated, including:
When dispatching non-leaf nodes correspondence calculating task operation reduction preset duration in topological diagram, the meter is recalculated
The end time of running of calculation task;
It is activation node to set the non-leaf nodes;
Judge it is described activation node correspondence calculating task the end time of running, if influence it is described activation node it is direct under
Bringing into operation the moment for node correspondence calculating task is swum, the direct downstream node is in the scheduling topological diagram, with swashing
The downstream node that movable joint point is joined directly together;
If influence, it is activation node to determine the direct downstream node, and recalculates activation node correspondence calculating task
Bring into operation the moment and terminate the time of running;
Judge whether the activation node is leaf node;
If it is, determining that the activation node is the leaf node that is activated;
If not, returning to circulation performs the end time of running for calculating the activation node correspondence calculating task, if influence
It is described activation node direct downstream node correspondence calculating task bring into operation the moment the step for.
11. method according to claim 10, it is characterised in that calculate the activation node correspondence calculating task
Terminate the time of running, if moment when bringing into operation of the direct downstream node correspondence calculating task of the influence activation node,
Including:
Scheduling topological diagram in, it is determined that activation node direct downstream node, and, the direct downstream node it is direct
Upstream node;
In all nodes immediately upstream of a direct downstream node, it is determined that the end time of running of correspondence calculating task the latest;
Whether the end time of running described in judging the latest is identical with activating the end time of running that node is recalculated;
If identical, it is determined that activation node influences bringing into operation the moment for direct downstream node correspondence calculating task, otherwise,
It is determined that activation node does not influence bringing into operation the moment for direct downstream node correspondence calculating task.
12. any one method according to claim 9-11, it is characterised in that methods described also includes:
All non-leaf nodes in topological diagram will be dispatched, be ranked up by system bottleneck degree size;
By the non-leaf nodes of predetermined number, system bottleneck node is used as.
13. a kind of device for determining upstream node bottleneck degree, it is characterised in that including:
Upstream node module is determined, all upstream nodes for determining present node;
It is determined that time range module can be optimized, for according to upstream node correspondence calculating task bring into operation the moment and
Present node corresponds to bringing into operation the moment for calculating task, it is determined that can optimize time range;
Module is split, for time range can be optimized, multiple time intervals are split as;
Upstream node bottleneck degree computing module, for according to each upstream node correspondence calculating task run time interval,
Time interval runs the total number of calculating task and can optimize time range, calculates the bottleneck degree of each upstream node.
14. device according to claim 13, it is characterised in that it is determined that time range module can be optimized, including:
Timing calculating task judging unit, for judging to whether there is timing calculating task in upstream node correspondence calculating task;
First object brings into operation moment determining unit, for when there is timing calculating task, the timing being calculated and is appointed
Bringing into operation the moment for business, brings into operation the moment as target;
Second target brings into operation moment determining unit, for when in the absence of timing calculating task, in upstream node,
Calculating task bringing into operation the moment earliest is determined, is that target brings into operation the moment;
It is determined that time range unit can be optimized, for being brought into operation moment calculating task corresponding with present node according to target
Bring into operation the moment, it is determined that time range can be optimized.
15. the device according to claim 14 or 13, it is characterised in that described device also includes:
First order module, for all upstream nodes of present node to be ranked up by bottleneck degree size;
Upstream bottlenecks node determining module, for by the upstream node of predetermined number, being used as the upstream bottlenecks section of present node
Point.
16. the device according to claim 13 or 14, it is characterised in that described device also includes:
Indirect bottleneck interval determination unit, for will have multiple calculating tasks while the time interval run, is used as indirect bottle
Between neck region;
Optimize difficulty determining unit, for according to indirect bottleneck is interval and optimizable time range, determining present node
Optimization difficulty.
17. a kind of device for determining system bottleneck degree, it is characterised in that including:
Shortest path computing module, the shortest path for a non-leaf nodes in calculating scheduling topological diagram to each leaf node
Footpath, the non-leaf nodes dispatches the node that there is downstream node in topological diagram to be described;
Leaf node operation duration determining module, for when non-leaf nodes correspondence calculating task increase operation is default
When long, according to shortest path, it is determined that the increased duration of calculating task operation institute corresponding to each leaf node;
The first system bottleneck degree determining module, for according to the increased duration of each leaf node operation institute, determining described non-
The system bottleneck degree of leaf node.
18. device according to claim 17, it is characterised in that shortest path computing module, including:
All path determining units, in scheduling topological diagram, determining all of a non-leaf nodes a to leaf node
Path;Wherein, each path includes multiple nodes;
Path length determining unit, for according in each path adjacent node correspondence calculating task the end time of running and
Bring into operation the moment, it is determined that the path length in each path;
Shortest path determining unit, is shortest path for determining path length most short path.
19. device according to claim 17, it is characterised in that leaf node operation duration determining module, including:
First judging unit, for judging present node correspondence calculating task, whether the increased preset duration of operation institute is more than
Present node to a leaf node shortest path length;
First operation duration determining unit, for present node correspondence calculating task, the increased preset duration of operation institute is more than
Present node to a leaf node shortest path length when, determine that leaf node correspondence calculating task operation is increased
Shi Changwei:Present node correspondence calculating task operation increased preset duration-shortest path length;
Second operation duration determining unit, for present node correspondence calculating task, the increased preset duration of operation institute is less than
Equal to present node to a leaf node shortest path length when, determine leaf node correspondence calculating task operation institute
A length of zero when increased.
20. the device according to any one of claim 17-19, it is characterised in that described device also includes:
Second order module, for all non-leaf nodes that will be dispatched in topological diagram, is arranged by system bottleneck degree size
Sequence;
The first system bottleneck determining module, for by the non-leaf nodes of predetermined number, being used as system bottleneck node.
21. a kind of device for determining system bottleneck degree, it is characterised in that including:
Non-leaf nodes determining module, for determining a non-leaf nodes;
Calculating is activated leaf node number module, for when non-leaf nodes correspondence calculating task operation reduce it is default
During duration, the number for the leaf node that is activated in scheduling topological diagram is calculated, the non-leaf nodes is the scheduling topological diagram
The middle node that there is downstream node, the leaf node that is activated is bring into operation moment and the end fortune of correspondence calculating task
The leaf node that the row moment changes;
Computing system bottleneck degree module, for the number and preset duration of the leaf node that is activated according to, determines institute
State the system bottleneck degree of non-leaf nodes.
22. device according to claim 21, it is characterised in that calculate the leaf node number module that is activated, bag
Include:
First recalculates unit, for default when dispatching non-leaf nodes correspondence calculating task operation reduction in topological diagram
During duration, the end time of running of the calculating task is recalculated;
Setting unit, for setting the non-leaf nodes to be activation node;
Judging unit is influenceed, the end time of running for judging the activation node correspondence calculating task, if influence institute
The direct downstream node for stating activation node corresponds to bringing into operation the moment for calculating task, and the direct downstream node is described
Dispatch in topological diagram, the downstream node being joined directly together with activation node;
Node determining unit is activated, if for influenceing, it is activation node to determine the direct downstream node, and is counted again
Calculate bringing into operation moment and the end time of running for activation node correspondence calculating task;
Leaf node determining unit, for judging whether the activation node is leaf node;
Be activated leaf node determining unit, for when it is leaf node to activate node, determining that the activation node is quilt
Activate leaf node.
23. device according to claim 22, it is characterised in that influence judging unit, including:
Direct downstream and immediately upstream determination subelement, in scheduling topological diagram, it is determined that the direct downstream of activation node
Node, and, the node immediately upstream of the direct downstream node;
Terminate time of running determination subelement the latest, in all nodes immediately upstream of a direct downstream node, really
Surely the end time of running of calculating task the latest is corresponded to;
Identical judgment sub-unit, for judging that the end that the end time of running the latest is recalculated with activation node is transported
Whether the row moment is identical;
Influence determining unit, for ought be the latest the ends time of running with activating the end time of running phase that recalculates of node
Meanwhile, it is determined that activation node influences bringing into operation the moment for direct downstream node correspondence calculating task;
Determining unit is not influenceed, the end time of running recalculated for the end time of running that ought be the latest and activation node
When differing, it is determined that activation node does not influence bringing into operation the moment for direct downstream node correspondence calculating task.
24. any one device according to claim 21-23, it is characterised in that described device also includes:
3rd order module, for all non-leaf nodes that will be dispatched in topological diagram, is arranged by system bottleneck degree size
Sequence;
Second system bottleneck determining module, for by the non-leaf nodes of predetermined number, being used as system bottleneck node.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610197115.3A CN107291533B (en) | 2016-03-31 | 2016-03-31 | Method and device for determining upstream node bottleneck degree and system bottleneck degree |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610197115.3A CN107291533B (en) | 2016-03-31 | 2016-03-31 | Method and device for determining upstream node bottleneck degree and system bottleneck degree |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107291533A true CN107291533A (en) | 2017-10-24 |
CN107291533B CN107291533B (en) | 2020-10-30 |
Family
ID=60086749
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610197115.3A Active CN107291533B (en) | 2016-03-31 | 2016-03-31 | Method and device for determining upstream node bottleneck degree and system bottleneck degree |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107291533B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111309712A (en) * | 2020-03-16 | 2020-06-19 | 北京三快在线科技有限公司 | Optimized task scheduling method, device, equipment and medium based on data warehouse |
CN113918298A (en) * | 2021-10-12 | 2022-01-11 | 北京字节跳动网络技术有限公司 | Method, system and equipment for processing multimedia data |
CN114500318A (en) * | 2022-02-11 | 2022-05-13 | 中国工商银行股份有限公司 | Batch operation monitoring method and device, equipment and medium |
CN114510329A (en) * | 2022-01-21 | 2022-05-17 | 北京火山引擎科技有限公司 | Method, device and equipment for determining predicted output time of task node |
CN115134224A (en) * | 2022-08-31 | 2022-09-30 | 杭州玳数科技有限公司 | DAG graph monitoring method and system |
CN115640988A (en) * | 2022-11-18 | 2023-01-24 | 北京大学 | Emergency event processing method and device and computer readable storage medium |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102201965A (en) * | 2010-03-24 | 2011-09-28 | 北京创世网赢高科技有限公司 | Data transmission method |
CN102378382A (en) * | 2010-08-10 | 2012-03-14 | 华为技术有限公司 | Method, equipment and system for scheduling data streams |
CN102467532A (en) * | 2010-11-12 | 2012-05-23 | 中国移动通信集团山东有限公司 | Task processing method and task processing device |
CN103593288A (en) * | 2013-11-06 | 2014-02-19 | 东北师范大学 | Test method for WS-BPEL workflow induction QoS analysis |
CN104503833A (en) * | 2014-12-22 | 2015-04-08 | 广州唯品会网络技术有限公司 | Task scheduling optimization method and device |
CN104778076A (en) * | 2015-04-27 | 2015-07-15 | 东南大学 | Scheduling method for cloud service workflow |
US20150296483A1 (en) * | 2011-01-14 | 2015-10-15 | Apple Inc. | Apparatus and methods for optimizing scheduled operations in hybrid network environments |
-
2016
- 2016-03-31 CN CN201610197115.3A patent/CN107291533B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102201965A (en) * | 2010-03-24 | 2011-09-28 | 北京创世网赢高科技有限公司 | Data transmission method |
CN102378382A (en) * | 2010-08-10 | 2012-03-14 | 华为技术有限公司 | Method, equipment and system for scheduling data streams |
CN102467532A (en) * | 2010-11-12 | 2012-05-23 | 中国移动通信集团山东有限公司 | Task processing method and task processing device |
US20150296483A1 (en) * | 2011-01-14 | 2015-10-15 | Apple Inc. | Apparatus and methods for optimizing scheduled operations in hybrid network environments |
CN103593288A (en) * | 2013-11-06 | 2014-02-19 | 东北师范大学 | Test method for WS-BPEL workflow induction QoS analysis |
CN104503833A (en) * | 2014-12-22 | 2015-04-08 | 广州唯品会网络技术有限公司 | Task scheduling optimization method and device |
CN104778076A (en) * | 2015-04-27 | 2015-07-15 | 东南大学 | Scheduling method for cloud service workflow |
Non-Patent Citations (1)
Title |
---|
王爱民: "《制造执行系统(MES)实现原理与技术》", 31 March 2014 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111309712A (en) * | 2020-03-16 | 2020-06-19 | 北京三快在线科技有限公司 | Optimized task scheduling method, device, equipment and medium based on data warehouse |
CN113918298A (en) * | 2021-10-12 | 2022-01-11 | 北京字节跳动网络技术有限公司 | Method, system and equipment for processing multimedia data |
CN114510329A (en) * | 2022-01-21 | 2022-05-17 | 北京火山引擎科技有限公司 | Method, device and equipment for determining predicted output time of task node |
CN114510329B (en) * | 2022-01-21 | 2023-08-08 | 北京火山引擎科技有限公司 | Method, device and equipment for determining estimated output time of task node |
CN114500318A (en) * | 2022-02-11 | 2022-05-13 | 中国工商银行股份有限公司 | Batch operation monitoring method and device, equipment and medium |
CN114500318B (en) * | 2022-02-11 | 2024-02-06 | 中国工商银行股份有限公司 | Batch operation monitoring method, device, equipment and medium |
CN115134224A (en) * | 2022-08-31 | 2022-09-30 | 杭州玳数科技有限公司 | DAG graph monitoring method and system |
CN115640988A (en) * | 2022-11-18 | 2023-01-24 | 北京大学 | Emergency event processing method and device and computer readable storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN107291533B (en) | 2020-10-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107291533A (en) | Determine method, the device of upstream node bottleneck degree and system bottleneck degree | |
CN106982139B (en) | Method and device for assigning problems based on multi-robot multitask and user equipment | |
CN106156082B (en) | A kind of ontology alignment schemes and device | |
CN108090093A (en) | The method and apparatus for generating recommendation results | |
Kiritsis et al. | Petri net techniques for process planning cost estimation | |
CN102155942A (en) | Global path planning method based on fuzzy topological map under large-scale environment | |
CN107861974B (en) | A kind of adaptive network crawler system and its data capture method | |
CN108009642A (en) | Distributed machines learning method and system | |
CN104765729A (en) | Cross-platform micro-blogging community account matching method | |
CN106845630A (en) | Unordered process must be through the shortest path acquisition methods and device of point | |
CN103793349A (en) | Data processing method and data processing device | |
CN108401424A (en) | A kind of the determination method, apparatus and robot of stacking path mode | |
CN108776686A (en) | Data tag construction system and method | |
CN108259195A (en) | The determining method and system of the coverage of anomalous event | |
CN106323324B (en) | Quick shortest path planning method based on road chain | |
CN109947948A (en) | A kind of knowledge mapping expression learning method and system based on tensor | |
CN109144514A (en) | JSON formatted data parses storage method and device | |
CN107291746A (en) | A kind of method and apparatus for storing and reading data | |
CN103294791A (en) | Extensible markup language pattern matching method | |
CN105894179A (en) | Business state transition method and system based on dynamic planning | |
CN102737134B (en) | Query processing method being suitable for large-scale real-time data stream | |
CN101882259A (en) | Method and equipment for filtering entity relationship instance | |
CN104239314B (en) | A kind of method and system of query expansion word | |
CN101297290A (en) | Method for controlling relational database system | |
CN105184373A (en) | Bayesian network structure learning method and system and reliability model construction method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |