CN106713164A - System, method and apparatus for balancing flow between switch ports - Google Patents
System, method and apparatus for balancing flow between switch ports Download PDFInfo
- Publication number
- CN106713164A CN106713164A CN201611250318.0A CN201611250318A CN106713164A CN 106713164 A CN106713164 A CN 106713164A CN 201611250318 A CN201611250318 A CN 201611250318A CN 106713164 A CN106713164 A CN 106713164A
- Authority
- CN
- China
- Prior art keywords
- cluster
- load
- affiliated
- excessive
- service
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/12—Avoiding congestion; Recovering from congestion
- H04L47/125—Avoiding congestion; Recovering from congestion by balancing the load, e.g. traffic engineering
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention provides a system, method and apparatus for balancing flow between switch ports. Each cluster connected with a switch comprises a first LB, and the first LBs of the clusters communicate with each other. Each first LB monitors whether the load of the belonging cluster is too large, if so, transfers one service provided by the belonging cluster to other clusters. The service is dynamically transferred, the flow balance between different clusters connected with the switch is realized, namely, the flow balance between ports of the switch is realized, and thus the resource utilization rate is improved.
Description
Technical field
The present invention relates to switch technology field, more specifically to flow equalization system between a kind of switch ports themselves,
Method and device.
Background technology
At present, four-layer load-equalizing equipment upstream side is carried out in cluster (Cluster) generally by the way of OSPF/ECMP
Flow equalization between portion's multiple LB (Load Balancer, load equalizer).Topological structure is as shown in figure 1, a upstream exchanges
The multiple clusters of machine (Switch) connection, each cluster externally provides multiple services (Service).Under this pattern, same exchange
Machine connection different clusters between uninterrupted also differ, easily cause between each port of interchanger flow difference compared with
Greatly, so cause resource utilization it is relatively low.If for example, the flow of Service 1 and Service 2 is much larger than Service 5
Flow, there is the bottleneck right side relatively low situation of flow simultaneously in the left side port for easily causing interchanger.
The content of the invention
In view of this, flow equalization system, method and device between the present invention proposes a kind of switch ports themselves, are intended to realize exchanging
Flow equalization between generator terminal mouthful, and then improve the purpose of resource utilization.
To achieve these goals, it is proposed that scheme it is as follows:
Flow equalization system between a kind of switch ports themselves, including:The multiple clusters being connected with interchanger, each described cluster
Including a LB (Load Balancer, load equalizer);
Whether each described LB, the load for monitoring its affiliated cluster is excessive, if so, its affiliated cluster is provided
A service be transferred to target cluster, the target cluster is to be selected from cluster other described.
Preferably, whether the load for monitoring its affiliated cluster is excessive, including:
The percentage load of its affiliated cluster is calculated, judges the percentage load whether more than default first threshold
Value, if so, then confirming the load excessive of its affiliated cluster.
Preferably, whether the load for monitoring its affiliated cluster is excessive, including:
The standard deviation of the percentage load of the multiple cluster is calculated, judges the standard deviation whether more than default the
Two threshold values, if so, whether the percentage load for then judging its affiliated cluster is maximum in the multiple cluster, if, it is determined that
The load excessive of its affiliated cluster.
Preferably, each described LB, is additionally operable to that the service that its affiliated cluster is provided is transferred into mesh described
Before mark cluster, whether the time of judging distance its preceding transfer service exceedes default very first time threshold value, if so, being then transferred to institute
State the service that its affiliated cluster is provided and be transferred to target cluster step;
And/or,
Each described LB, before being additionally operable to receive the service of other clusters transfer, judging distance once receives before it turn
Whether the time of shifting service exceedes default second time threshold, if so, then receive the transfer of this service, if it is not, then refusing this
The transfer of secondary service.
Flow equalization method between a kind of switch ports themselves, is applied to a LB, and methods described includes:
Whether the load of cluster is excessive belonging to monitoring;
If so, the service that affiliated cluster is provided then is transferred into target cluster, said target cluster is from other collection
Selected in group.
Preferably, whether the load of cluster belonging to the monitoring is excessive, including:
It is calculated the percentage load of affiliated cluster;
Judge the percentage load whether more than preset first threshold value;
If so, the load excessive of cluster belonging to then confirming.
Preferably, whether the load of cluster belonging to the monitoring is excessive, including:
It is calculated the standard deviation of the percentage load of all clusters being connected with interchanger;
Judge the standard deviation whether more than default Second Threshold;
If the standard deviation is more than default Second Threshold, whether the percentage load for judging affiliated cluster is described all
It is maximum in cluster;
If the percentage load of affiliated cluster is maximum in all clusters, it is determined that the load excessive of affiliated cluster.
Flow equalization device between a kind of switch ports themselves, is applied to a LB, and described device includes:
Whether monitoring unit, the load for monitoring affiliated cluster is excessive;
Buanch unit, during for load excessive in affiliated cluster, mesh is transferred to by the service that affiliated cluster is provided
Mark cluster, said target cluster is to be selected from other clusters.
Preferably, the monitoring unit, including:
First computation subunit, the percentage load for being calculated affiliated cluster;
First judgment sub-unit, for judging the percentage load whether more than preset first threshold value;
First confirms subelement, for when the percentage load is more than preset first threshold value, confirming affiliated cluster
Load excessive.
Preferably, the monitoring unit, including:
Second computation subunit, the standard of the percentage load for being calculated all clusters being connected with interchanger
Difference;
Second judgment sub-unit, for judging the standard deviation whether more than default Second Threshold;
3rd judgment sub-unit, if being more than default Second Threshold for the standard deviation, judges the load of affiliated cluster
Whether percentage is maximum in all clusters;
Second confirms subelement, if being maximum in all clusters for the percentage load of affiliated cluster, it is determined that
The load excessive of affiliated cluster.
Compared with prior art, technical scheme has advantages below:
Flow equalization system, method and device between a kind of switch ports themselves that above-mentioned technical proposal is provided.Connect with interchanger
Each cluster for connecing includes a LB, is communicated between a LB of each cluster.Each LB monitors collection belonging to it
Whether the load of group is excessive, if so, the service that its affiliated cluster is provided then is transferred into other clusters.Dynamically taken
Business transfer, reaches the flow equalization between the different clusters of interchanger connection, that is, realize the flow equalization between each port of interchanger, enters
And improve resource utilization.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
The accompanying drawing to be used needed for having technology description is briefly described, it should be apparent that, drawings in the following description are only this
Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can be with
Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is the principle schematic of load balancing;
Fig. 2 is the load balancing schematic diagram between each cluster of interchanger connection provided in an embodiment of the present invention;
Fig. 3 is the structural representation of flow equalization system between a kind of switch ports themselves provided in an embodiment of the present invention;
Fig. 4 is the flow chart of flow equalization method between a kind of switch ports themselves provided in an embodiment of the present invention;
Fig. 5 is the flow chart of flow equalization method between another switch ports themselves provided in an embodiment of the present invention;
Fig. 6 is the flow chart of flow equalization method between another switch ports themselves provided in an embodiment of the present invention;
Fig. 7 is the structural representation of flow equalization device between a kind of switch ports themselves provided in an embodiment of the present invention;
Fig. 8 is the structural representation of flow equalization device between another switch ports themselves provided in an embodiment of the present invention;
Fig. 9 is the structural representation of flow equalization device between another switch ports themselves provided in an embodiment of the present invention.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.It is based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made
Embodiment, belongs to the scope of protection of the invention.
The present embodiment provides flow equalization system between a kind of switch ports themselves, and referring to shown in Fig. 2 and Fig. 3, the system includes:
The multiple clusters 2 being connected with interchanger 1, each described cluster 2 includes a LB, phase between a LB of each cluster 2
Mutual communication, realizes the information sharing of each cluster.
Whether each described LB, the load for monitoring its affiliated cluster is excessive;
For whether the load for judging cluster is excessive, a LB of each cluster calculates the percentage load of this cluster.And bear
Carry percentage calculation, can with preset using configure by the way of be configured to equipment execution logic in, if each
First LB of cluster uses unified computational methods.Specific calculation those skilled in the art can be according to actual need
Carry out free setting.For ease of the understanding to percentage load, the scheme of being implemented as described below is the embodiment of the invention provides, but need
To illustrate that it is the unique implementation of embodiment that following scheme is not the present invention, should not be construed as to the embodiment of the present invention
Uniqueness is limited, and specific percentage load η is defined as follows:
Wherein, LlbnRepresent the present flow rate speed of n-th LB and switchboard direct connector in cluster, ClbnIn expression cluster
The permission maximum stream flow speed of n-th LB and switchboard direct connector, WlbnRepresent n-th data weighting of LB in cluster.
Each described LB, is additionally operable in the load excessive of its affiliated cluster, that its affiliated cluster is provided
Service is transferred to target cluster, and the target cluster is to be selected from cluster other described.
For one suitable target cluster of selection carries out service transfer, a LB of each cluster calculates the appearance of this cluster
Amount.And the calculation of capacity, can with preset using configuration by the way of be configured to equipment execution logic in, if each
First LB of cluster uses unified computational methods.Specific calculation those skilled in the art can be according to actual need
Carry out free setting.For ease of the understanding to percentage load, the scheme of being implemented as described below is the embodiment of the invention provides, but need
To illustrate that it is the unique implementation of embodiment that following scheme is not the present invention, should not be construed as to the embodiment of the present invention
Uniqueness is limited, specific capacity CclusterIt is defined as follows:
Ccluster=Wlb1×Clb1+Wlb2×Clb2+…+Wlbn×Clbn
Wherein, ClbnRepresent the permission maximum stream flow speed of n-th LB and switchboard direct connector in cluster, WlbnRepresent collection
N-th data weighting of LB in group.
The residual capacity C of each clusterPotential, can be according to its percentage load η and capacity CclusterObtain,
CPotential=Ccluster×(1-η).First LB of certain cluster, in the load excessive of its affiliated cluster, from other clusters
One residual capacity C of selectionPotentialMaximum or residual capacity CPotentialMore than predetermined threshold value cluster as target cluster.
First LB of certain cluster, when the service that its affiliated cluster is provided is transferred into target cluster, selects institute
Which service of category cluster carries out transfer those skilled in the art and can according to actual needs preset corresponding rule, this hair
Bright embodiment can be limited not, and this is because the technical scheme that the embodiment of the present invention is used is a process for dynamic adjustment.
If but selection uninterrupted is shifted more than the service of the residual capacity of target cluster herein, target cluster is will result in
Overload, then just there is no need to carry out service to displaced;Or selection uninterrupted is much smaller than the residual capacity of target cluster
Service shifted, it is as big as possible and turned less than the service of the residual capacity of target cluster that effect can be worse than selection flow
Move.Therefore, following preferred embodiment is the embodiment of the invention provides, it is specific as follows:To the service in this cluster according to stream
Amount size sorts according to this, whether its residual capacity for being more than target cluster is judged according to this from big to small, if it is not, then will be less than target
The service of the residual capacity of cluster is transferred to target cluster, and stops judging.
Flow equalization system between a kind of switch ports themselves that the present embodiment is provided.Each cluster being connected with interchanger is wrapped
Include a LB, the intercommunication of a LB of each cluster.Each LB monitor its affiliated cluster load whether
It is excessive, if so, the service that its affiliated cluster is provided then is transferred into other clusters.Service transfer dynamically is carried out, is reached
Flow equalization between the different clusters of interchanger connection, that is, realize the flow equalization between each port of interchanger, and then improve resource
Utilization rate.It is controlled without the interchanger to different manufacturers model, you can realize the flow equalization between switch ports themselves.
Specifically, a LB of each cluster monitor its affiliated cluster load whether excessive process includes:Calculate
To the percentage load of its affiliated cluster, whether the percentage load is judged more than preset first threshold value, if so, then confirming it
The load excessive of affiliated cluster.
The computing formula of the percentage load η of cluster can also be:
η=W1×ηCPU+W2×ηNIC+W3×ηMemory
Wherein, ηCPURepresent the cpu load percentage of cluster, W1Represent the weight of cpu data, ηNICRepresent that cluster network interface card is born
Carry percentage, W2Represent the weight of cluster network card data, ηMemoryRepresent cluster memory percentage load, W3Represent internal storage data
Weight;Llb1nRepresent n-th CPU use values, the C of LB in clusterlb1nRepresent that n-th CPU of LB can be used maximum, WlbnRepresent
N-th data weighting of LB, L in clusterlb2nThe present flow rate speed of n-th LB and switchboard direct connector in cluster is represented,
Clb2nRepresent the permission maximum stream flow speed of n-th LB and switchboard direct connector in cluster, Llb3nRepresent n-th LB in cluster
Internal memory use value, Clb3nN-th internal memory of LB can be used maximum in representing cluster.
The capacity C of corresponding clusterclusterComputing formula it is as follows:
Ccluster=W1×CCPU+W2×CNIC+W3×CMemory
CCPU=Wlb1×Clb11+Wlb2×Clb12+…+Wlbn×Clb1n
CNIC=Wlb1×Clb21+Wlb2×Clb22+…+Wlbn×Clb2n
CMemory=Wlb1×Clb31+Wlb2×Clb32+…+Wlbn×Clb3n
Wherein, CCPURepresent the CPU capacity of cluster, W1Represent the weight of cpu data, CNICRepresent cluster network interface card capacity, W2Table
Show the weight of cluster network card data, CMemoryRepresent cluster memory capacity, W3Represent the weight of internal storage data;Clb1nRepresent n-th
The CPU of LB can be used maximum, WlbnRepresent n-th data weighting of LB, C in clusterlb2nRepresent n-th LB and friendship in cluster
Change planes the permission maximum stream flow speed of direct-connected interface, Clb3nN-th internal memory of LB can be used maximum in representing cluster.
Another kind judges that whether excessive mode is for load, and the load difference degree of different clusters is calculated using standard deviation, when
When diversity factor (standard deviation) exceedes certain threshold values, the load represented between cluster is very unbalanced;Pressure is now chosen again
Migration or balance that larger cluster is serviced.During the average load of computing cluster, can average;In view of collection group energy
The different situation of power, it would however also be possible to employ weighted average is loaded.Specifically, a LB of each cluster monitors the negative of its affiliated cluster
Carry whether excessive process includes:The standard deviation of the percentage load of the multiple cluster is calculated, the standard deviation is judged
Whether more than default Second Threshold, if so, then judge its affiliated cluster percentage load whether be in the multiple cluster most
Greatly, if, it is determined that the load excessive of its affiliated cluster.
For the connection for avoiding frequent migration (transfer) service from causing is unstable, each cluster can set and migrate twice most
The closely-spaced time.The double minimum time interval for receiving migration can also be set.Specifically, a LB of each cluster,
Be additionally operable to it is described the service that its affiliated cluster is provided is transferred to target cluster before, judging distance is once shifted before it and taken
Whether the time of business exceedes default very first time threshold value, if so, being then transferred to a service turn for providing its affiliated cluster
Move to target cluster step;
And/or,
Each described LB, before being additionally operable to receive the service of other clusters transfer, judging distance once receives before it turn
Whether the time of shifting service exceedes default second time threshold, if so, then receive the transfer of this service, if it is not, then refusing this
The transfer of secondary service.
That is could be arranged to 1) each cluster minimum interval of transfer service, 2) receiving of each cluster twice
The minimum interval of transfer service, 3) each cluster the minimum interval of transfer service and receive transfer service twice twice
Minimum interval.
The present embodiment provides flow equalization method between a kind of switch ports themselves, is applied to a LB of above-described embodiment, joins
See Fig. 4, the method includes:
Step S11:Whether the load of cluster is excessive belonging to monitoring;
Step S12:If so, the service that affiliated cluster is provided then is transferred into target cluster, said target cluster is
Selected from other clusters.
The present embodiment provide a kind of switch ports themselves between flow equalization method, monitoring belonging to cluster load whether mistake
Greatly, if so, the service that affiliated cluster is provided then is transferred into other clusters.Service transfer dynamically is carried out, exchange is reached
Flow equalization between the different clusters of machine connection, that is, realize the flow equalization between each port of interchanger, and then improve the utilization of resources
Rate.It is controlled without the interchanger to different manufacturers model, you can realize the flow equalization between switch ports themselves.
The present embodiment also provide cluster belonging to a kind of monitoring load whether excessive specific method, referring to described in Fig. 5, should
Method includes:
Step S111:It is calculated the percentage load of affiliated cluster;
Step S112:Judge the percentage load whether more than preset first threshold value;
Step S113:If so, the load excessive of cluster belonging to then confirming.
The present embodiment also provide cluster belonging to another monitoring load whether excessive specific method, referring to described in Fig. 6,
The method includes:
Step S114:It is calculated the standard deviation of the percentage load of all clusters being connected with interchanger;
Step S115:Judge the standard deviation whether more than default Second Threshold;
Step S116:If the standard deviation is more than default Second Threshold, whether the percentage load of cluster belonging to judgement
It is maximum in all clusters;
Step S117:If the percentage load of affiliated cluster is maximum in all clusters, it is determined that affiliated cluster
Load excessive.
Specifically, when a LB judges the load excessive of its affiliated cluster, the process for carrying out services migrating is as follows:
1st, the maximum cluster of residual capacity is chosen as target cluster;
2nd, in some services that this cluster is provided, selection uninterrupted is as big as possible, and surplus no more than target cluster
The service of covolume amount;
3rd, mass-sended services migrating to object set to ask, request contains the information on services to be migrated;
If the 4, receiving " reception " of target cluster return, illustrate that target cluster agrees to migration, if receiving what is returned
Message is " refusal ", then continue to select suboptimum cluster, and continue to attempt to;
5th, after being agreed migration, the configuration of the service to be migrated is carried out, target cluster passes through after the completion of configuration is migrated
OSPF (Open Shortest Path First, ospf) externally announces the VIP of the service to be migrated
(Virtual IP Address, virtual ip address), ECMP (Equal-CostMultipathRouting, equal cost multipath) meeting
Guarantee service is imported into target cluster.
6th, after services migrating, former cluster stops externally announcing the VIP of the service to be migrated, and updates the negative of the cluster
Carry percentage and capacity and externally bulletin (after target cluster is finished receiving to the service for migrating, also updates own load percentage
Announce with capacity and externally).
The present embodiment also provides flow equalization device between a kind of switch ports themselves, is applied to a LB, shown in Figure 7,
Described device includes:
Whether monitoring unit 11, the load for monitoring affiliated cluster is excessive;
Buanch unit 12, during for load excessive in affiliated cluster, the service that affiliated cluster is provided is transferred to
Target cluster, said target cluster is to be selected from other clusters.
Flow equalization device between a kind of switch ports themselves that the present embodiment is provided, detection unit 11, cluster belonging to monitoring
Whether load is excessive;The service that affiliated cluster is provided is transferred to other by buanch unit in the load excessive of the cluster
Cluster.Service transfer dynamically is carried out, the flow equalization between the different clusters of interchanger connection is reached, that is, realizes that interchanger is respectively held
Flow equalization between mouthful, and then improve resource utilization.It is controlled without the interchanger to different manufacturers model, you can realize
Flow equalization between switch ports themselves.
Present embodiments provide a kind of concrete structure of monitoring unit, shown in Figure 8, the monitoring unit 11, specifically
Including:
First computation subunit 111, the percentage load for being calculated affiliated cluster;
First judgment sub-unit 112, for judging the percentage load whether more than preset first threshold value;
First confirms subelement 113, for when the percentage load is more than preset first threshold value, confirming affiliated cluster
Load excessive.
Present embodiments provide a kind of concrete structure of monitoring unit, shown in Figure 9, the monitoring unit 11, specifically
Including:
Second computation subunit 114, the mark of the percentage load for being calculated all clusters being connected with interchanger
It is accurate poor;
Second judgment sub-unit 115, for judging the standard deviation whether more than default Second Threshold;
3rd judgment sub-unit 116, if being more than default Second Threshold for the standard deviation, judges the negative of affiliated cluster
Carry whether percentage is maximum in all clusters;
Second confirms subelement 117, if being maximum in all clusters for the percentage load of affiliated cluster, really
The load excessive of cluster belonging to fixed
For device embodiment, because it essentially corresponds to embodiment of the method, so related part is referring to method reality
Apply the part explanation of example.Device embodiment described above is only schematical, wherein described as separating component
The unit of explanation can be or may not be physically separate, and the part shown as unit can be or can also
It is not physical location, you can with positioned at a place, or can also be distributed on multiple NEs.Can be according to reality
Selection some or all of module therein is needed to realize the purpose of this embodiment scheme.Those of ordinary skill in the art are not
In the case of paying creative work, you can to understand and implement.
Herein, such as first and second or the like relational terms be used merely to by an entity or operation with it is another
One entity or operation make a distinction, and not necessarily require or imply these entities or there is any this reality between operating
Relation or order.And, term " including ", "comprising" or its any other variant be intended to the bag of nonexcludability
Contain, so that process, method, article or equipment including a series of key elements are not only including those key elements, but also including
Other key elements being not expressly set out, or it is this process, method, article or the intrinsic key element of equipment also to include.
In the absence of more restrictions, the key element limited by sentence "including a ...", it is not excluded that including the key element
Process, method, article or equipment in also there is other identical element.
Each embodiment is described by the way of progressive in this specification, and what each embodiment was stressed is and other
The difference of embodiment, between each embodiment identical similar portion mutually referring to.
To the described above of disclosed embodiment of this invention, professional and technical personnel in the field is realized or use this
Invention.Various modifications to these embodiments will be apparent for those skilled in the art, institute herein
The General Principle of definition can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore,
The present invention is not intended to be limited to the embodiments shown herein, and is to fit to special with principles disclosed herein and novelty
The consistent scope most wide of point.
Claims (10)
1. flow equalization system between a kind of switch ports themselves, it is characterised in that including:The multiple clusters being connected with interchanger, often
The individual cluster includes first load equalizer;
Whether each described first load equalizer, the load for monitoring its affiliated cluster is excessive, if so, by its affiliated cluster
The service for providing is transferred to target cluster, and the target cluster is to be selected from cluster other described.
2. system according to claim 1, it is characterised in that whether the load of the monitoring its affiliated cluster is excessive, bag
Include:
Whether the percentage load of its affiliated cluster is calculated, the percentage load is judged more than preset first threshold value, if
It is then to confirm the load excessive of its affiliated cluster.
3. system according to claim 1, it is characterised in that whether the load of the monitoring its affiliated cluster is excessive, bag
Include:
The standard deviation of the percentage load of the multiple cluster is calculated, judges the standard deviation whether more than default second threshold
Value, if so, whether the percentage load for then judging its affiliated cluster is maximum in the multiple cluster, if, it is determined that its institute
Belong to the load excessive of cluster.
4. the system according to claims 1 to 3 any one, it is characterised in that each described first load equalizer, also
For it is described the service that its affiliated cluster is provided is transferred to target cluster before, judging distance its preceding transfer service
Time whether exceed default very first time threshold value, if so, being then transferred to a service transfer for providing its affiliated cluster
To target cluster step;
And/or,
Each described first load equalizer, before being additionally operable to receive the service of other clusters transfer, judging distance is once connect before it
Whether exceeded default second time threshold by the time of transfer service, if so, then receive the transfer of this service, if it is not, then refusing
The transfer of this service absolutely.
5. flow equalization method between a kind of switch ports themselves, it is characterised in that be applied to the first load equalizer, methods described bag
Include:
Whether the load of cluster is excessive belonging to monitoring;
If so, the service that affiliated cluster is provided then is transferred into target cluster, said target cluster is from other clusters
Selection.
6. method according to claim 5, it is characterised in that whether the load of cluster is excessive belonging to the monitoring, including:
It is calculated the percentage load of affiliated cluster;
Judge the percentage load whether more than preset first threshold value;
If so, the load excessive of cluster belonging to then confirming.
7. method according to claim 5, it is characterised in that whether the load of cluster is excessive belonging to the monitoring, including:
It is calculated the standard deviation of the percentage load of all clusters being connected with interchanger;
Judge the standard deviation whether more than default Second Threshold;
If the standard deviation is more than default Second Threshold, whether the percentage load for judging affiliated cluster is all clusters
Middle maximum;
If the percentage load of affiliated cluster is maximum in all clusters, it is determined that the load excessive of affiliated cluster.
8. flow equalization device between a kind of switch ports themselves, it is characterised in that be applied to the first load equalizer, described device bag
Include:
Whether monitoring unit, the load for monitoring affiliated cluster is excessive;
Buanch unit, during for load excessive in affiliated cluster, object set is transferred to by the service that affiliated cluster is provided
Group, said target cluster is to be selected from other clusters.
9. device according to claim 8, it is characterised in that the monitoring unit, including:
First computation subunit, the percentage load for being calculated affiliated cluster;
First judgment sub-unit, for judging the percentage load whether more than preset first threshold value;
First confirms subelement, for when the percentage load is more than preset first threshold value, confirming the load of affiliated cluster
It is excessive.
10. device according to claim 8, it is characterised in that the monitoring unit, including:
Second computation subunit, the standard deviation of the percentage load for being calculated all clusters being connected with interchanger;
Second judgment sub-unit, for judging the standard deviation whether more than default Second Threshold;
3rd judgment sub-unit, if being more than default Second Threshold for the standard deviation, judges the load percentage of affiliated cluster
Than whether be in all clusters it is maximum;
Second confirms subelement, if being maximum in all clusters for the percentage load of affiliated cluster, it is determined that affiliated
The load excessive of cluster.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611250318.0A CN106713164A (en) | 2016-12-29 | 2016-12-29 | System, method and apparatus for balancing flow between switch ports |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611250318.0A CN106713164A (en) | 2016-12-29 | 2016-12-29 | System, method and apparatus for balancing flow between switch ports |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106713164A true CN106713164A (en) | 2017-05-24 |
Family
ID=58904058
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611250318.0A Pending CN106713164A (en) | 2016-12-29 | 2016-12-29 | System, method and apparatus for balancing flow between switch ports |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106713164A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107197478A (en) * | 2017-07-19 | 2017-09-22 | 中国人民解放军理工大学 | A kind of flow equalization method based on forwarding ionization meter |
CN107707612A (en) * | 2017-08-10 | 2018-02-16 | 北京奇艺世纪科技有限公司 | A kind of appraisal procedure and device of the resource utilization of load balancing cluster |
CN108282526A (en) * | 2018-01-22 | 2018-07-13 | 中国软件与技术服务股份有限公司 | Server dynamic allocation method and system between double clusters |
CN111464602A (en) * | 2020-03-24 | 2020-07-28 | 平安银行股份有限公司 | Flow processing method and device, computer equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101778002A (en) * | 2010-02-02 | 2010-07-14 | 浪潮(北京)电子信息产业有限公司 | Large-scale cluster system and building method thereof |
CN103916450A (en) * | 2013-12-10 | 2014-07-09 | 中国海洋石油总公司 | Foundation architecture of high-performance data warehouse |
CN104243337A (en) * | 2013-06-09 | 2014-12-24 | 杭州华三通信技术有限公司 | Method and device for cross-cluster load balancing |
CN105282191A (en) * | 2014-06-20 | 2016-01-27 | 中国电信股份有限公司 | Load balancing system, controller and method |
-
2016
- 2016-12-29 CN CN201611250318.0A patent/CN106713164A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101778002A (en) * | 2010-02-02 | 2010-07-14 | 浪潮(北京)电子信息产业有限公司 | Large-scale cluster system and building method thereof |
CN104243337A (en) * | 2013-06-09 | 2014-12-24 | 杭州华三通信技术有限公司 | Method and device for cross-cluster load balancing |
CN103916450A (en) * | 2013-12-10 | 2014-07-09 | 中国海洋石油总公司 | Foundation architecture of high-performance data warehouse |
CN105282191A (en) * | 2014-06-20 | 2016-01-27 | 中国电信股份有限公司 | Load balancing system, controller and method |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107197478A (en) * | 2017-07-19 | 2017-09-22 | 中国人民解放军理工大学 | A kind of flow equalization method based on forwarding ionization meter |
CN107707612A (en) * | 2017-08-10 | 2018-02-16 | 北京奇艺世纪科技有限公司 | A kind of appraisal procedure and device of the resource utilization of load balancing cluster |
CN108282526A (en) * | 2018-01-22 | 2018-07-13 | 中国软件与技术服务股份有限公司 | Server dynamic allocation method and system between double clusters |
CN111464602A (en) * | 2020-03-24 | 2020-07-28 | 平安银行股份有限公司 | Flow processing method and device, computer equipment and storage medium |
CN111464602B (en) * | 2020-03-24 | 2023-04-18 | 平安银行股份有限公司 | Flow processing method and device, computer equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110618870B (en) | Working method and device for deep learning training task | |
CN106713164A (en) | System, method and apparatus for balancing flow between switch ports | |
CN103560967B (en) | The virtual data center mapping method of a kind of business demand perception | |
CN107347205B (en) | A kind of network slice selection method, apparatus and system | |
CN103385033B (en) | For the method and apparatus distributing session bundle in NE | |
CN104158755B (en) | The methods, devices and systems of transmitting message | |
CN109660466A (en) | A kind of more live load balance realizing methods towards cloud data center tenant | |
CN104092621B (en) | A kind of load sharing method and device | |
CN106332182A (en) | Wireless access method and apparatus | |
CN102164081B (en) | Routing computation method of fat-tree topology, node equipment and communication system | |
CN102110014A (en) | Method for balancing loads of virtual machine (VM) | |
CN104852859B (en) | A kind of aggregation interface method for processing business and equipment | |
CN100426778C (en) | Adaptive polling method of router | |
CN101217528A (en) | A link aggregation method and device | |
CN105162897A (en) | System and method for allocating IP address for virtual machine and network virtual machine | |
CN106776014A (en) | Parallel acceleration method and system in Heterogeneous Computing | |
CN109905280A (en) | A kind of emulation mode and system towards mobile satellite network | |
CN104022951B (en) | A kind of method for building up and system in network service path | |
CN106817306A (en) | A kind of method and device for determining target route | |
CN115174404B (en) | Multi-device federal learning system based on SDN networking | |
CN106453121B (en) | A kind of link dynamic load configuration method, system and server | |
CN106059940A (en) | Flow control method and device | |
CN105704180B (en) | The configuration method and its system of data center network | |
CN106451415A (en) | Microgrid power transaction system based on distributed coordination method | |
CN107835130A (en) | A kind of flow allocation method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170524 |