CN104579975A - Method for dispatching software-defined network controller cluster - Google Patents
Method for dispatching software-defined network controller cluster Download PDFInfo
- Publication number
- CN104579975A CN104579975A CN201510073823.1A CN201510073823A CN104579975A CN 104579975 A CN104579975 A CN 104579975A CN 201510073823 A CN201510073823 A CN 201510073823A CN 104579975 A CN104579975 A CN 104579975A
- Authority
- CN
- China
- Prior art keywords
- controller node
- controller
- switch
- data
- node
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Computer And Data Communications (AREA)
Abstract
The invention relates to a method for dispatching a software-defined network controller cluster. According to the method, a scheme of a controller cluster is adopted to allocate loading pressure to each controller node, so that the problem of relatively low processing performance due to the fact that all loading pressure is intensively given to a single controller node can be avoided; a Hash table is correspondingly configured to each controller node in the cluster, and each controller node is only permitted to modify the identification and load factors thereof of a switch which is in the Hash table and is taken over by that controller node; when data in the Hash table of each controller node is modified by the respective controller node by each controller node, the modified data is synchronized to other controller nodes, and data in other Hash tables is correspondingly updated by respective controller nodes; and corresponding controller nodes are determined according to data in the Hash tables to take over the switch, a second-party manager does not need to be introduced, and the performance of the whole cluster is not limited to the third-party manager.
Description
Technical field
The present invention relates to networking technology area, be specifically related to a kind of dispatching method of software defined network controller cluster.
Background technology
At present, software defined network (Software Defined Network, SDN) scheduling scheme of controller roughly comprises two kinds, a kind of is scheduling scheme based on two-node cluster hot backup, its framework as shown in Figure 1, be responsible for taking over all switches by master controller node at ordinary times, when master controller node occurs delaying machine, take over all switches by preparation controller node; Its shortcoming is that load pressure all concentrates on single controller node, and handling property is lower.
Another kind is the scheduling scheme being carried out controller cluster by third party manager, and its framework as shown in Figure 2, adds a third party manager outside all controller nodes, is in charge of all controller nodes in scheduling cluster; All controller nodes are all the taking over tasks being distributed switch by this third party manager; If the machine of delaying appears in controller node, third party manager will reschedule switch taking over tasks; Its shortcoming is if the machine of delaying appears in third party manager, the operation of overall network system can be affected, meanwhile, because the number of nodes upper limit of controller cluster is by the gerentocratic program scheduler performance impact of third party, the gerentocratic performance of third party is made can to become the bottleneck of overall cluster.
Summary of the invention
The present invention is directed to the scheduling scheme handling property of software defined network controller in prior art lower or be limited to the technical problems such as the gerentocratic performance of third party, propose a kind of dispatching method with more high performance software defined network controller cluster.
The dispatching method of a kind of software defined network controller cluster that the present invention proposes, described software defined network controller cluster comprises some controller nodes; Each controller node is corresponding to be had a Hash and shows, and the data of every part of Hash table comprise: the mark of each controller node, each controller node are responsible for the mark of switch, the load factor of each controller node taken over; Wherein, each controller node only allows to revise mark and its load factor of being responsible for the switch taken in its Hash table had by it; When each controller node have modified the data in Hash table that it has, the data syn-chronization that need revise in other controller nodes, the data during the Hash had separately by the corresponding renewal of other controller nodes shows;
Described dispatching method comprises controller node electoral machinery, and described controller node electoral machinery comprises:
This controller node obtains the mark of the switch needing election adapter;
According to the data in the Hash table that this controller node has, judge whether the switch taking over the mark correspondence got.
Preferably, the data during the described Hash had according to this controller node shows, judge whether that the step of the switch taking over the mark correspondence got comprises:
According to the data in the Hash table that this controller node has, judge whether the switch of the mark correspondence got is taken over by other controller nodes;
If so, then this controller node abandons adapter; If not, then the switch quantity of taking over according to this controller node judges whether adapter.
Preferably, described switch quantity of taking over according to this controller node judges whether that the step of taking over comprises:
According to the data in the Hash table that this controller node has, judge that whether the switch quantity that this controller node is taken over is minimum; If so, then judge whether to take over according to the switch quantity whether having other controller nodes to take over is equal with the switch quantity that this controller node is taken over.
Preferably, the described basis switch quantity that whether has other controller nodes to take over is equal with the switch quantity that this controller node is taken over judges whether that the step of taking over comprises:
According to the data in the Hash table that this controller node has, judge whether that the switch quantity that other controller nodes are taken over is equal with the switch quantity that this controller node is taken over;
If not, then this controller node is taken over; If so, then adapter is judged whether according to the identification number of this controller node.
Preferably, the described identification number according to this controller node judges whether that the step of taking over comprises:
Data in the Hash table had according to this controller node, judge whether the identification number of this controller node is minimum in the identification number of all adapter switch quantity controller node equal with it;
If so, then this controller node is taken over.
Preferably, the switch of this controller node to the mark correspondence got also comprises after taking over:
This controller node revises mark and its load factor of being responsible for the switch taken in its Hash table had by it, and the data syn-chronization revised is in other controller nodes; Data in the Hash table had separately by the corresponding renewal of other controller nodes.
Preferably, the data syn-chronization revised especially by TCP of this controller node is in other controller nodes.
Preferably, this controller node sends request the packet of adapter to switch by openflow agreement, to realize the adapter of the switch to the mark correspondence got.
Preferably, described dispatching method also comprises synchrodata restorative procedure, and described synchrodata restorative procedure comprises:
Inquire about in the Hash table that this controller node has and whether there is abnormal data, the data in the Hash table if so, then had this controller node are repaired, and in the Hash table simultaneously had other controller nodes, corresponding data are synchronously repaired.
Preferably, described dispatching method also comprises load-balancing method, and described load-balancing method comprises:
Data in the Hash table had according to this controller node judge whether this controller node is in load idle condition, if so, the switch of the then initiatively controller node that takeover load is busy.
Beneficial effect: the dispatching method of the software defined network controller cluster that the present invention proposes, the scheme of what it adopted is controller cluster, load pressure can be shared to each controller node, avoid in the scheduling scheme of prior art two-node cluster hot backup and all load pressures are concentrated the problem causing handling property lower to single controller node;
Simultaneously, the a Hash table of each controller node corresponding configuration in cluster of the present invention, and limit each controller node and only allow to revise mark and its load factor of being responsible for the switch taken in its Hash table had by it, when each controller node have modified the data in its Hash table had, the data syn-chronization that need be revised in other controller nodes, the data in the Hash that had separately by the corresponding renewal of other controller nodes table; Therefore, according to the data in Hash table, the present invention can determine that corresponding controller node is taken over switch, and do not need to introduce third party manager, so whole cluster can not be caused to be limited to the gerentocratic performance of third party.
Show based on described Hash, the present invention can realize redundancy between each controller node and load balancing, substantially increases performance and the efficiency of whole cluster.
Accompanying drawing explanation
Fig. 1 is the configuration diagram of two-node cluster hot backup in prior art.
Fig. 2 is the configuration diagram of being carried out controller cluster in prior art by third party manager.
Fig. 3 is the configuration diagram of the dispatching method embodiment of a kind of software defined network controller cluster that the present invention proposes.
Fig. 4 is the overall flow figure of the dispatching method embodiment middle controller node electoral machinery of a kind of software defined network controller cluster that the present invention proposes.
Embodiment
For the ease of it will be appreciated by those skilled in the art that the present invention is described further below in conjunction with accompanying drawing and embodiment.
The dispatching method embodiment of a kind of software defined network controller cluster that the present invention proposes, refer to Fig. 3, described software defined network controller cluster comprises some controller nodes (controller node is at least more than 2); Each controller node is corresponding to be had a Hash and shows, and the data of every part of Hash table comprise: the mark of each controller node, each controller node are responsible for the mark of switch, the load factor of each controller node taken over; Wherein, each controller node only allows to revise mark and its load factor of being responsible for the switch taken in its Hash table had by it; When each controller node have modified the data in Hash table that it has, the data syn-chronization that need revise in other controller nodes, the data during the Hash had separately by the corresponding renewal of other controller nodes shows.
The form of the Hash table in the present embodiment is as shown in table 1:
Table 1
| The identification number of controller node | The mark of current adapter switch | Load factor |
| NodeId 1 | swId1 | 1 |
| NodeId Id 2 | SwId4、swId5、swId6 | 3 |
| NodeId Id 3 | swId7、swId8 | 2 |
| NodeId Id 4 | Empty | 0 |
| ... | ... | ... |
In Table 1, NodeId 1 to NodeId Id 4 is the identification number of each controller node, the mark of swId1 to be identification number the be switch of the current adapter of controller node of NodeId1; Can find out in table 1, identification number is the load factor (be also weight, can be expressed as weight) of the controller node of NodeId 1 is 1 simultaneously.
In the present embodiment, the form of the preferred adoption rate of load factor, so that load balancing computing.Such as, as can be seen from Table 1, identification number is the controller node of NodeId 1 current adapter 1 switch (swId1), and its load factor is 1; Identification number is the current adapter of controller node 3 switches (SwId4, swId5, swId6) of NodeId 2, and its load factor is 3; Identification number is the current adapter of controller node 2 switches (swId7, swId8) of NodeId 3, and its load factor is 2; The load factor of this three is 1:3:2.Meanwhile, in order to make the performance of whole cluster better, the value of load factor preferably controls within 1 to 10.
Other symbols related in table 1 can carry out corresponding understanding according to above-mentioned explanation.
Hash table in the present embodiment can realize the schemes such as controller node election, synchrodata reparation and load balancing in the cluster.The present embodiment introduces the scheme of controller node election.
Concrete, in the present embodiment, the dispatching method of software defined network controller cluster comprises controller node electoral machinery, and described controller node electoral machinery entirety comprises step S100 and step S200:
S100, this controller node obtain the mark of the switch needing election adapter;
Data in S200, the Hash table that has according to this controller node, judge whether the switch taking over the mark correspondence got.
In the present embodiment, described controller node electoral machinery is mainly used in two kinds of situations: one is, when there being new switch to be connected to this cluster, each controller node in this cluster can receive response, and each controller node receiving response can run described controller node selection algorithm; Two are, when have controller node delay machine time, its adapter switch need by election taken over by other normal controller nodes, other normal controller nodes determine whether to take over based on described controller node selection algorithm.
Data during the Hash that controller node electoral machinery in the present embodiment mainly configures based on each controller node shows realize, each controller node can realize by TCP the synchronized update that Hash shows data, simultaneously, each controller node can send request the packet of adapter to switch by openflow agreement, to realize the adapter of the switch to the mark correspondence got.
In order to realize the object of the present embodiment better, below described controller node electoral machinery being done and optimizing further and introduce in detail.
In the present embodiment, step S200 specifically comprises step S210 and step S220:
Data in S210, the Hash table that has according to this controller node, judge whether the switch of the mark correspondence got is taken over by other controller nodes;
S220, if so, then this controller node abandon adapter; If not, then the switch quantity of taking over according to this controller node judges whether adapter.
Described Hash table, for table 1, if the mark that step S100 gets is any one in table 1 in swId1 to swId8, then illustrates that the switch of this mark correspondence is taken over by controller node.Such as, suppose that the mark got is " swId8 " here, the identification number of this controller node is " NodeId Id 2 ", the switch that adapter is designated " swId8 " is not had by 1 known controller node of tabling look-up, the just identified sequence number of switch being designated " swId8 " is taken over by the controller of " NodeId Id 3 ", and therefore this controller node abandons adapter; If the mark that step S100 gets is " swId9 ", known by tabling look-up, the switch being designated " swId9 " is not taken over by any controller node, and therefore, this controller node can judge whether adapter according to the switch quantity of self current adapter.
From shown in table 1, identification number is that the switch quantity of the controller node adapter of NodeId 1 to NodeId 4 is respectively 1,3,2 and 0.
In the present embodiment, in order to prevent switch from losing, only when switch is taken over by other controller nodes, just this controller node can abandon the adapter to this switch, otherwise electoral machinery can be performed, until this switch is taken over always; But when judging that switch is not taken over by other controller nodes, this controller node may not necessarily directly be taken over.Preferably, this controller node can in conjunction with self-condition, judges it self is whether currently meet most the controller node carrying out taking over, takes over corresponding determining whether.Therefore, the present embodiment is further optimized on the basis of step S220.
Concrete, the switch quantity of taking over according to this controller node in step S220 judges whether that the step of taking over specifically comprises step S230:
Data in S230, the Hash table that has according to this controller node, judge that whether the switch quantity that this controller node is taken over is minimum; If so, then judge whether to take over according to the switch quantity whether having other controller nodes to take over is equal with the switch quantity that this controller node is taken over.
Simultaneously, step S230 also can optimize further, concrete, the switch quantity that in step S230, whether basis has other controller nodes to take over is equal with the switch quantity that this controller node is taken over judges whether that the step of taking over comprises step S240 and step S250:
Data in S240, the Hash table that has according to this controller node, judge whether that the switch quantity that other controller nodes are taken over is equal with the switch quantity that this controller node is taken over;
S250, if not, then this controller node is taken over; If so, then adapter is judged whether according to the identification number of this controller node.
The Main Function of above-mentioned steps S230 to step S250 avoids occurring between each controller node that adapter is disorder (such as, occur that multiple controller node is taken over simultaneously or all do not take over the situation of certain switch), and the adapter rule set out, to improve the standardized management to controller node.
In the present embodiment, in step S250, judge whether that the step of taking over specifically comprises step S260 and step S270 according to the identification number of this controller node:
Data in S260, the Hash table that has according to this controller node, judge whether the identification number of this controller node is minimum in the identification number of all adapter switch quantity controller node equal with it;
S270, if so, then this controller node is taken over.
Step S260 to step S270, mainly in order to do further standardized management to controller node, namely first using the controller node that identification number is less, re-using the controller node that identification number is larger, facilitate user to the application of controller node and management.Certainly, also can be obviously other modes in the present embodiment, such as, first use the controller node that identification number is larger, re-use the controller node that identification number is less.
In the present embodiment, the switch of this controller node to the mark correspondence got also comprises step S300 after taking over:
S300, this controller node revise mark and its load factor of being responsible for the switch taken in its Hash table had by it, and the data syn-chronization revised is in other controller nodes; Data in the Hash table had separately by the corresponding renewal of other controller nodes.
Overall flow after above step S100 to step S300 integrates refers to Fig. 4.
In the preferred version of the present embodiment, except described controller node electoral machinery, the dispatching method of a kind of software defined network controller cluster that the present embodiment proposes also comprises synchrodata restorative procedure, and described synchrodata restorative procedure comprises:
Inquire about in the Hash table that this controller node has and whether there is abnormal data, the data in the Hash table if so, then had this controller node are repaired, and in the Hash table simultaneously had other controller nodes, corresponding data are synchronously repaired.
In the present embodiment, each controller node can be currently connected to the mark of all switches in this cluster by openflow agreement query.Showing by traveling through this locality (this controller node) Hash, inquiring about a switch and being taken over by how many controller nodes, if having and only have the adapter of controller node, illustrate and be in normal condition to switch; If do not have controller node to take over this switch, this controller node will take over this switch by openflow agreement in time, and inform other controller nodes.If there is two or more controller node to take over this switch, illustrate wherein one or more controller nodes data exist abnormal; If this controller node is included in wherein, this controller node needs to send Query Information by openflow agreement, timely reparation or upgrade the data of the table of Hash on this controller node, in the Hash table that has of other controller nodes, corresponding data are synchronously repaired or upgrade simultaneously.
In the present embodiment further preferred scheme, except described controller node electoral machinery, synchrodata restorative procedure, the dispatching method of a kind of software defined network controller cluster that the present embodiment proposes also comprises synchrodata restorative procedure and also comprises load-balancing method, and described load-balancing method comprises:
Judge whether this controller node is in load idle condition according to the data in the Hash table that this controller node has, if, the then initiatively switch abandoned of the needs of the controller node that takeover load is busy, until this controller node reaches load saturation state.
In the present embodiment, whether each controller node timing self can be in load idle condition, if be in load idle condition, then continuing to calculate also needs adapter how many switches just can reach load saturation state, and then go to calculate other controller nodes and whether be in load busy state, if it is busy that other controller nodes are in load, it needs to abandon within one or more switch makes self to be in load saturation state, (load idle condition) this controller node controller node that initiatively takeover load is busy can need the switch abandoned (this controller node first takes over this switch, reinform the adapter that former controller node is abandoned this switch), until load saturation state.Here adapter work can not upgrade the data in the Hash table that this controller node has, and more the work of new data can complete in synchrodata restorative procedure.
In sum, the dispatching method of a kind of software defined network controller cluster that the present embodiment proposes, the scheme of controller cluster due to what adopt, load pressure can be shared to each controller node, avoid in the scheduling scheme of prior art two-node cluster hot backup and all load pressures are concentrated the problem causing handling property lower to single controller node.The a Hash table of each controller node corresponding configuration in the cluster of the present embodiment, and limit each controller node and only allow to revise mark and its load factor of being responsible for the switch taken in its Hash table had by it, when each controller node have modified the data in its Hash table had, the data syn-chronization that need be revised in other controller nodes, the data in the Hash that had separately by the corresponding renewal of other controller nodes table; Therefore, according to the data in Hash table, the present embodiment can determine that corresponding controller node is taken over switch, and do not need to introduce third party manager, so whole cluster can not be caused to be limited to the gerentocratic performance of third party, thus greatly improve efficiency and the performance of overall cluster.
The above embodiment only have expressed several execution mode of the present invention, and it describes comparatively concrete and detailed, but therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection range of patent of the present invention should be as the criterion with claims.
Claims (10)
1. a dispatching method for software defined network controller cluster, described software defined network controller cluster comprises some controller nodes; It is characterized in that, each controller node is corresponding to be had a Hash and shows, and the data of every part of Hash table comprise: the mark of each controller node, each controller node are responsible for the mark of switch, the load factor of each controller node taken over; Wherein, each controller node only allows to revise mark and its load factor of being responsible for the switch taken in its Hash table had by it; When each controller node have modified the data in Hash table that it has, the data syn-chronization that need revise in other controller nodes, the data during the Hash had separately by the corresponding renewal of other controller nodes shows;
Described dispatching method comprises controller node electoral machinery, and described controller node electoral machinery comprises:
This controller node obtains the mark of the switch needing election adapter;
According to the data in the Hash table that this controller node has, judge whether the switch taking over the mark correspondence got.
2. the dispatching method of software defined network controller cluster according to claim 1, is characterized in that, the data during the described Hash had according to this controller node shows, judge whether that the step of the switch taking over the mark correspondence got comprises:
According to the data in the Hash table that this controller node has, judge whether the switch of the mark correspondence got is taken over by other controller nodes;
If so, then this controller node abandons adapter; If not, then the switch quantity of taking over according to this controller node judges whether adapter.
3. the dispatching method of software defined network controller cluster according to claim 2, is characterized in that, described switch quantity of taking over according to this controller node judges whether that the step of taking over comprises:
According to the data in the Hash table that this controller node has, judge that whether the switch quantity that this controller node is taken over is minimum; If so, then judge whether to take over according to the switch quantity whether having other controller nodes to take over is equal with the switch quantity that this controller node is taken over.
4. the dispatching method of software defined network controller cluster according to claim 3, it is characterized in that, the switch quantity whether described basis has other controller nodes to take over is equal with the switch quantity that this controller node is taken over judges whether that the step of taking over comprises:
According to the data in the Hash table that this controller node has, judge whether that the switch quantity that other controller nodes are taken over is equal with the switch quantity that this controller node is taken over;
If not, then this controller node is taken over; If so, then adapter is judged whether according to the identification number of this controller node.
5. the dispatching method of software defined network controller cluster according to claim 4, is characterized in that, the described identification number according to this controller node judges whether that the step of taking over comprises:
Data in the Hash table had according to this controller node, judge whether the identification number of this controller node is minimum in the identification number of all adapter switch quantity controller node equal with it;
If so, then this controller node is taken over.
6., according to the dispatching method of the arbitrary described software defined network controller cluster of claim 1 to 5, it is characterized in that, the switch of this controller node to the mark correspondence got also comprises after taking over:
This controller node revises mark and its load factor of being responsible for the switch taken in its Hash table had by it, and the data syn-chronization revised is in other controller nodes; Data in the Hash table had separately by the corresponding renewal of other controller nodes.
7. the dispatching method of software defined network controller cluster according to claim 6, is characterized in that, the data syn-chronization that this controller node is revised especially by TCP is in other controller nodes.
8. the dispatching method of software defined network controller cluster according to claim 7, it is characterized in that, this controller node sends request the packet of adapter to switch by openflow agreement, to realize the adapter of the switch to the mark correspondence got.
9. the dispatching method of software defined network controller cluster according to claim 8, it is characterized in that, described dispatching method also comprises synchrodata restorative procedure, and described synchrodata restorative procedure comprises:
Inquire about in the Hash table that this controller node has and whether there is abnormal data, the data in the Hash table if so, then had this controller node are repaired, and in the Hash table simultaneously had other controller nodes, corresponding data are synchronously repaired.
10. the dispatching method of software defined network controller cluster according to claim 9, it is characterized in that, described dispatching method also comprises load-balancing method, and described load-balancing method comprises:
Judge whether this controller node is in load idle condition according to the data in the Hash table that this controller node has, if, the then initiatively switch abandoned of the needs of the controller node that takeover load is busy, until this controller node reaches load saturation state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510073823.1A CN104579975B (en) | 2015-02-10 | 2015-02-10 | A kind of dispatching method of software defined network controller cluster |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510073823.1A CN104579975B (en) | 2015-02-10 | 2015-02-10 | A kind of dispatching method of software defined network controller cluster |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104579975A true CN104579975A (en) | 2015-04-29 |
| CN104579975B CN104579975B (en) | 2018-01-05 |
Family
ID=53095200
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510073823.1A Active CN104579975B (en) | 2015-02-10 | 2015-02-10 | A kind of dispatching method of software defined network controller cluster |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104579975B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109697070A (en) * | 2017-10-23 | 2019-04-30 | 中移(苏州)软件技术有限公司 | A kind of cluster management method based on Ambari, device and medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011065268A1 (en) * | 2009-11-26 | 2011-06-03 | 日本電気株式会社 | Load distribution system, load distribution method, and program |
| CN102594697A (en) * | 2012-02-21 | 2012-07-18 | 华为技术有限公司 | Load balancing method and device |
| US20130028091A1 (en) * | 2011-07-27 | 2013-01-31 | Nec Corporation | System for controlling switch devices, and device and method for controlling system configuration |
| CN102970379A (en) * | 2012-12-19 | 2013-03-13 | 中国电子科技集团公司第十五研究所 | Method for realizing load balance among multiple servers |
| CN103618621A (en) * | 2013-11-21 | 2014-03-05 | 华为技术有限公司 | Method, device and system for automatic configuration of SDN |
| CN103929333A (en) * | 2014-05-08 | 2014-07-16 | 陈桂芳 | Implementation method for SDN controller pool |
| CN104104572A (en) * | 2014-07-15 | 2014-10-15 | 杭州华三通信技术有限公司 | Automatic deploying method and device for switch in SDN network |
| CN104301243A (en) * | 2014-09-22 | 2015-01-21 | 华为技术有限公司 | Load control method and device |
-
2015
- 2015-02-10 CN CN201510073823.1A patent/CN104579975B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011065268A1 (en) * | 2009-11-26 | 2011-06-03 | 日本電気株式会社 | Load distribution system, load distribution method, and program |
| US20130028091A1 (en) * | 2011-07-27 | 2013-01-31 | Nec Corporation | System for controlling switch devices, and device and method for controlling system configuration |
| CN102594697A (en) * | 2012-02-21 | 2012-07-18 | 华为技术有限公司 | Load balancing method and device |
| CN102970379A (en) * | 2012-12-19 | 2013-03-13 | 中国电子科技集团公司第十五研究所 | Method for realizing load balance among multiple servers |
| CN103618621A (en) * | 2013-11-21 | 2014-03-05 | 华为技术有限公司 | Method, device and system for automatic configuration of SDN |
| CN103929333A (en) * | 2014-05-08 | 2014-07-16 | 陈桂芳 | Implementation method for SDN controller pool |
| CN104104572A (en) * | 2014-07-15 | 2014-10-15 | 杭州华三通信技术有限公司 | Automatic deploying method and device for switch in SDN network |
| CN104301243A (en) * | 2014-09-22 | 2015-01-21 | 华为技术有限公司 | Load control method and device |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109697070A (en) * | 2017-10-23 | 2019-04-30 | 中移(苏州)软件技术有限公司 | A kind of cluster management method based on Ambari, device and medium |
| CN109697070B (en) * | 2017-10-23 | 2022-02-18 | 中移(苏州)软件技术有限公司 | Ambari-based cluster management method, device and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104579975B (en) | 2018-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112867050B (en) | UPF network element management method and system | |
| US9154382B2 (en) | Information processing system | |
| CN108965014B (en) | QoS-aware service chain backup method and system | |
| CN108076158B (en) | Minimum load route selection method and system based on naive Bayes classifier | |
| CN108206790B (en) | SDN joint routing selection and resource allocation method based on network slice | |
| CN107729147A (en) | Data processing method, control node and stream calculation system in stream calculation system | |
| CN104579801A (en) | Method for dispatching software-defined network controller cluster | |
| CN104753828A (en) | SDN controller, data central system and route connection method | |
| CN107615709A (en) | SDN retransmission unit and controller unit | |
| EP3214800A1 (en) | Method and device for implementing capacity planning | |
| CN103891086A (en) | Providing electrical energy | |
| WO2016197301A1 (en) | Policy coordination method and apparatus in nfv system | |
| CN104539551A (en) | Virtual service migration method for routing and switching platform | |
| JP2014225872A (en) | Method for mapping network topology request to physical network, computer program product, mobile communication system and network configuration platform | |
| US10216593B2 (en) | Distributed processing system for use in application migration | |
| CN105721328A (en) | VRRP load balancing method, VRRP load balancing device and router | |
| US11700189B2 (en) | Method for performing task processing on common service entity, common service entity, apparatus and medium for task processing | |
| CN103795606A (en) | Service transfer control method based on sdn in optical network | |
| Sridharan et al. | Multi-controller traffic engineering in software defined networks | |
| CN104579975A (en) | Method for dispatching software-defined network controller cluster | |
| JP5950021B2 (en) | Control device, communication system, synchronization method and program | |
| JP6619708B2 (en) | Data separation evaluation apparatus, data separation evaluation method, and data separation evaluation program | |
| WO2019019915A1 (en) | Scheduling solution configuration method and apparatus, computer readable storage medium thereof, and computer device | |
| CN110851186B (en) | Network equipment restarting method and device, electronic equipment and readable storage medium | |
| CN109218179A (en) | Load sharing method and device, interchanger and computer readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 510000 building G1, No. 17, software Road, Guangzhou, Guangdong, Tianhe District Applicant after: Guangzhou hi tech software Limited by Share Ltd Address before: 510000 building G1, No. 17, software Road, Guangzhou, Guangdong, Tianhe District Applicant before: Guangzhou Pingao Software Development Co., Ltd. |
|
| COR | Change of bibliographic data | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |