CN103685461B - A kind of cluster management device, management system and management method - Google Patents

A kind of cluster management device, management system and management method Download PDF

Info

Publication number
CN103685461B
CN103685461B CN201310507575.8A CN201310507575A CN103685461B CN 103685461 B CN103685461 B CN 103685461B CN 201310507575 A CN201310507575 A CN 201310507575A CN 103685461 B CN103685461 B CN 103685461B
Authority
CN
China
Prior art keywords
server node
request
external system
session
session management
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.)
Expired - Fee Related
Application number
CN201310507575.8A
Other languages
Chinese (zh)
Other versions
CN103685461A (en
Inventor
王金铨
黄代良
殴哲毅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Telefonaktiebolaget LM Ericsson AB
Original Assignee
Telefonaktiebolaget LM Ericsson AB
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Telefonaktiebolaget LM Ericsson AB filed Critical Telefonaktiebolaget LM Ericsson AB
Priority to CN201310507575.8A priority Critical patent/CN103685461B/en
Publication of CN103685461A publication Critical patent/CN103685461A/en
Application granted granted Critical
Publication of CN103685461B publication Critical patent/CN103685461B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention discloses a kind of cluster management device, management system and management method.Cluster management device provided in an embodiment of the present invention takes the mode of server cluster, the control of internal server node is issued for the request message selection of external system by session management module, selected according to the performance of the type of request message and server node, realize dynamic load balancing, therefore increase cluster server node when business is occurred frequently and usually reduce clustered node, cost of investment has been saved, has reduced the wasting of resources, while do not interfere with service operation.

Description

A kind of cluster management device, management system and management method
Technical field
The present invention relates to communication technical field, and in particular to a kind of cluster management device, management system and management method.
Background technology
The 3G charge systems of telecommunications industry typically have high steady using centralized server deployment operation, server more at present Qualitative, reliability, but as the development of 3G business, particularly data service are fast-developing, the prepayment of Real-time Billing System support Money number of users and set meal increases substantially at one's own expense, causes the middle-size and small-size server load for deploying charge system too high, from And in business high-incidence season meet Lingao risk, it is therefore desirable to constantly implement engineering to application and service device and be upgraded to more high-end clothes Business device, but the growth of the portfolio of centralized deployment needs that larger engineering must be implemented during dilatation, this can influence business and normally transport OK, and due to portfolio that communication service has time property concentrate it is frequently-occurring, as short message service amount is usually several times during the Spring Festival, Buying server needs to refer to traffic peak, therefore cost is high.
The content of the invention
The embodiment of the present invention provides a kind of cluster management device, management system and management method, in usually business Amount avoids the waste of resource when few, while and can meets the needs of when business is occurred frequently.
The embodiment of the present invention provides a kind of cluster management device, including:
Load balancing control module, for the proxy server in the request selecting flow-control module according to external system Node is issued;
Flow-control module, including more than one proxy server, for being attached license to the external system Control and overloading control;
Session management module, for the request of external system that is transmitted to the flow-control module internally server section Point establishes connection queue and creates session.
Wherein, the internal server node is set for bladeization, also, the session management module includes two or more Session management server.
Wherein, the flow-control module includes for being attached admissions control to external system:The flow control Module is attached authentication and protocol testing to the external system, when the connection authentication meets preparatory condition with agreement, The request of the external system is issued to the session management module.
In the first possible embodiment, with reference in a first aspect, the session management module is being received outside described After the request message of portion's system, the type of the request message is identified;The request is eliminated according to the type of the request message After noise in message;Request message selection internal server node is issued.
In second of possible embodiment, with reference to the possible embodiment party of the first of first aspect or first aspect Formula, the session management module establish the external system and internal server according to following factor more than one or two Connection queue between node:It is central processing unit model, network rate, memory rate, application process number, file system number, same When the compiler number that uses.
In the third possible embodiment, with reference in a first aspect, the session management module is controlled the flow Internally server node foundation connection queue includes for the request of the external system of module transmission:The session management module calculates The request message number of each internal server node and the difference of response message number, the minimum internal server node of the difference Corresponding balanced highest priority, the request of the external system is sent to the internal server section of balanced highest priority Point, if more than two inside points server nodes correspond to the minimum difference of identical, from the inside with identical minimum difference Selection one is carried out in server node at random, the internal server node as external system connection;Or
The session management module is according to the performance label value Y of each internal server noderCarry out the external system Request issues, Yr=Xr/min(X1, X2…Xr…Xy), wherein, XrFor the performance number of default r-th of internal server node, Min represents to take minimum value from the performance number of all internal server nodes, also, the property of each internal server node Energy value is set according to the average duration of each internal server node processing message.
In the 4th kind of possible embodiment, with reference in a first aspect, the session management module is additionally operable to establishment Session is monitored, when monitoring an internal server nodes break down corresponding to session, by the session equilibrium extremely Other internal server nodes, when monitoring that all internal server nodes all break down corresponding to session, the session Management module controls the flow-control module to enter off-line mode.
Wherein, it is connected with each other between flow-control module, session management module, load balancing control module.
Second aspect, the embodiment of the present invention also provide a kind of cluster management system, including:
More than one external system, asked for requested service processing;
More than one internal server node, for handling the business processing of the external system more than one Request;
Cluster management device, the cluster management device include load balancing control module, flow-control module and session Management module;
Wherein, load balancing control module, for the agency in the request selecting flow-control module according to external system Server node is issued;Flow-control module, including more than one proxy server, for entering to the external system Row connection admissions control and overloading control;Session management module, for the external system transmitted to the flow-control module Request internally server node establish connection queue and create session.
The third aspect, the embodiment of the present invention also provide a kind of cluster management method, including:
Receive the request from external system;
The proxy server nodes issued according to the request selecting of the external system received;
Request to the external system is attached admissions control and overloading control;
Internally server node establishes connection queue and creates session for request to the external system.
Cluster management device provided in an embodiment of the present invention uses server cluster, is external system by session management module Request message selection issue the control of internal server node, realize dynamic load balancing, increase cluster when business is occurred frequently Server node and usually reducing clustered node, saved cost of investment, reduced the wasting of resources, while do not interfered with business Operation.
Brief description of the drawings
Fig. 1 is the structural representation of cluster management device provided in an embodiment of the present invention;
Fig. 2 is the structural representation of another cluster management device provided in an embodiment of the present invention;
Fig. 3 is the DCC consultative management flow charts of session management module provided in an embodiment of the present invention;
Fig. 4 is the SMPP consultative management flow charts of session management module provided in an embodiment of the present invention;
Fig. 5 is the SGIP consultative management flow charts of session management module provided in an embodiment of the present invention;
Fig. 6 is the structural representation of cluster management system provided in an embodiment of the present invention;
Fig. 7 is the schematic flow sheet of cluster management method provided in an embodiment of the present invention.
Embodiment
The embodiment of the present invention provides a kind of cluster management device, management system and management method, in usually business Amount avoids the waste of resource when few, while and can meets the needs of when business is occurred frequently.
In order that those skilled in the art more fully understand the present invention program, below in conjunction with the embodiment of the present invention Accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only The embodiment of a part of the invention, rather than whole embodiments.Based on the embodiment in the present invention, ordinary skill people The every other embodiment that member is obtained under the premise of creative work is not made, it should all belong to the model that the present invention protects Enclose.
The embodiment of the present invention provides a kind of cluster management device, for realizing load balancing, including:
Load balancing control module, for the proxy server in the request selecting flow-control module according to external system Node is issued;
Flow-control module, including more than one proxy server, for being attached license to the external system Control and overloading control;
Session management module, for the request of external system that is transmitted to the flow-control module internally server section Point establishes connection queue and creates session.
Wherein, the internal server node is set for bladeization, also, the session management module can include two Session management server above, i.e. session management module take high availability cluster(High Available, check HA)Portion Management side formula, namely a set of internal server node can correspond to a session management server.
The establishment session refers to that it is logical that session management module establishes communication between external system and internal server node Road, after communication port is built up, the transmission of signal can be carried out between external system and internal server node, i.e., external system and Internal server node engages in the dialogue.
The flow-control module is attached admissions control to external system to be included:The flow-control module is to described External system is attached authentication and protocol testing, when the connection authentication meets preparatory condition with agreement, by the outside The request of system is issued to the session management module.
Further, for the session management module after the request message of the external system is received, identification is described please Seek the type of message;It is the request message after the noise in the request message is eliminated according to the type of the request message Selection internal server node is issued;
It should be noted that session management module not simply uses " dividing and rule " in the embodiment of the present invention, i.e., not That one big computational problem is turned to multiple small amount of calculation problems, be not the decomposition on problem but clustered node business processing The equilibrium of ability and the Dynamic Programming of portfolio, session management module can enter according to following factor more than one or two Row issues the control of internal server node:Central processing unit model, network rate, memory rate, application process number, file system System number, compiler number used at the same time, i.e. these factors can be as the comprehensive measurement index of clustered node disposal ability, root The Dynamic Programming of portfolio is realized according to this index.
In order to realize the Dynamic Programming of portfolio, selection of the session management module to issuing internal server node can root Under type is carried out according to this:
The session management module calculates the request message number of each internal server node and the difference of response message number, Balanced highest priority corresponding to the minimum internal server node of the difference, the request of the external system is sent to The internal server node for the highest priority that weighs, if more than two inside points server nodes correspond to the minimum difference of identical, Selection one is then carried out at random from the internal server node with identical minimum difference, as external system connection Internal server node;Or
The session management module is according to the performance label value Y of each internal server noderCarry out the external system Request issues, Yr=Xr/min(X1, X2…Xr…Xy), wherein, XrFor the performance number of default r-th of internal server node, Min represents to take minimum value from the performance number of all internal server nodes, also, the property of each internal server node Energy value is set according to the average duration of each internal server node processing message.
Further, the session that the session management module is additionally operable to establishment is monitored, when monitoring that session is corresponding An internal server nodes break down when, by the session equilibrium to other internal server nodes, when monitoring meeting When all internal server nodes all break down corresponding to words, the session management module controls the flow-control module to enter Enter off-line mode.
Specifically, cluster management device 100 shown in Figure 1, including:
Load balancing control module 101, for the agency in the request selecting flow-control module 102 according to external system Server node is issued;
Flow-control module 102, including more than one proxy server, for being attached perhaps to the external system Controllable and overloading control;
Session management module 103, the request of the external system for being transmitted to the flow-control module 102 internally take Business device node establishes connection queue and creates session.
Wherein, can be mutual between flow-control module 102, session management module 103 and load balancing control module 101 Connection;
Further, shown in Figure 2, in embodiments of the present invention, session management module 103 can include two or more Session management server, and take HA modes to dispose between more than two session management servers, flow-control module 102 Multiple proxy servers can be taken to dispose, flow-control module 102 is connected with session management module 103, i.e. session management module Each session management server in 103 is connected with each proxy server in flow-control module 102, flow-control module 102 send connection request to session management module 103, and session management module 103 services equivalent to flow-control module 102 End;
Further, in the embodiment of the present invention, internal server node is more than two servers, and internal server Node can take bladeization to dispose, i.e., be one group per N number of server node, wherein, N is positive integer, in fig. 2, every 3 clothes Device node be engaged in as one group, every group of internal server node corresponds to a session management server, in fig. 2, server node 1, Server node 2 and the respective session management server 1 of server node 3, server node 4, server node 5 and server section 6 respective session management servers 2 of point, server node 7, server node 8 and the respective session management server of server node 9 3。
It should be noted that in embodiments of the present invention, session management module 103 can conversate information concentrate it is standby Part, session routing iinformation statistics, overtime session management, application layer protocol management, node load balancing, malfunctioning node are again equal Weighing apparatus, clustered node management, man-machine interaction management;
Wherein, equilibrium refers to each internal server nodes break down to malfunctioning node again, can all be synchronized to session pipe Module 103 is managed, session management module 103 can be balanced to the session information for being routed to the malfunctioning node at other server nodes Reason, i.e., when session management server 1 monitors that server node 1 is faulty, session management server 1 can be server node Session information on 1 is transferred to server node 2 or server node 3 to handle;When session management module 103 monitors institute When having the internal server node all to break down, session management module 103 enters protected mode, and controls flow-control module 102 enter off-line mode.
Application layer protocol management includes DCC(Diameter Credit Control, diameter diameter credit control)Consultative management, SMPP(Short Message Peer to Peer, the transmission of short message point-to-point)Consultative management, SGIP(Short Message Gateway Interface Protocol, short message network management system interface)Consultative management, specific reference to Fig. 3, Fig. 4 and Fig. 5 It is shown;
Fig. 3 represents the flow of DCC consultative managements, in figure 3, the proxy server in flow-control module 102, for description Clear, it is DCC proxy servers that we, which define the proxy server, herein, actively initiates CER(Capability Exchange Request, capabilities exchange request)Connect the session management server in session management module 103;Session management server is replied CEA(Capability Exchange Acknowledge, capabilities exchange request-reply), establish DCC with proxy server and connect Connect;Proxy server sends CCR(Credit Control Request, credit control request)Message is to session management service Device;Session management server forwarding load balancing CCR to FEP(Front End Processor, front-end processor);Session pipe Manage server and reply CCA(Credit Control Acknowledge, diameter credit control are replied)To DCC proxy servers;Session Between management server and DCC proxy servers support DWR (Device Watchdog Request, house dog request message)/ DWA(Decice Watchdog Answer, house dog response message)Heartbeat packet;Session management server and DCC agency services DPR is also supported between device(Disconnect Peer Request, break link request)/DPA(Disconnect Peer Answer, break link response)Chain rupture message package;Session management server sends RAR(Re-authorization Request, re-authentication requests)Message is to DCC proxy servers;DCC proxy servers reply RRA(Re-authorization Answer, discrimination weight response)Message package.
Fig. 4 represents the flow of SMPP consultative managements, in Fig. 4, the proxy server in flow-control module 102, to retouch State clear, it is SMPP proxy servers that we, which define the proxy server, herein, actively initiates to receive to session management server Hold log on request bind_receiver;Session management server replys receiving terminal log on request and responds bind_receiver_ Resp, establish SMPP with SMPP proxy servers and be connected;SMPP proxy servers send withholding fee message auth_acc to session Management server;Session management server load balancing auth_acc to FEP is deducted fees;Session management server is replied and withheld Take response auth_acc_resp to SMPP proxy servers;Chain is supported between session management server and SMPP proxy servers Road maintains request enquire_link/ links to maintain response enquire_link_resp heartbeat packets;SMPP proxy servers are sent As a result notice asks result_notify to session management server;Session management server load balancing result_notify Extra money is carried out to FEP sides;Session management server replys result push-notification-answer result_notify_resp to SMPP agency's clothes Business device;Session management server supports that exiting request unbind/ exits response unbind_resp with SMPP proxy servers, from And exit connection.
Fig. 5 represents the flow of SGIP consultative managements, in Figure 5, to describe clear, the agency in flow-control module 102 Server is defined as SGIP proxy servers, actively initiates log on request bind connection session management servers;Session management takes Business device, which is replied, logs in response bind_resp, establishes SGIP with SMPP proxy servers and is connected;SGIP proxy servers send authentication Wholesale price asks SM_ACC message to session management server;Session management server load balancing SM_ACC to FEP is deducted fees; Session management server replys wholesale price request response SM_ACC_RESP to SGIP proxy servers;SGIP proxy servers are sent As a result notify result notice request SM_FM to session management server;Session management server load balancing SM_FM to FEP sides Carry out extra money;Session management server replys result push-notification-answer SM_FM_RESP to SGIP proxy servers;Session management takes Business device and SGIP proxy servers between support exit request unbind/ exit response unbind_resp exit connection.
Therefore cluster management device provided in an embodiment of the present invention takes the mode of server cluster, by session pipe Reason module issues the control of internal server node for the request message selection of external system, according to the type kimonos of request message The performance of business device node is selected, and realizes dynamic load balancing, thus increase when business is occurred frequently cluster server node and Clustered node is usually being reduced, is saving cost of investment, is reducing the wasting of resources, while do not interfere with service operation.
The embodiment of the present invention also provides a kind of cluster management system, shown in Figure 6 for realizing load balancing, including:
More than one external system, asked for requested service processing;
More than one internal server node, for handling the business processing of the external system more than one Request;
Cluster management device 100, the cluster management device include load balancing control module 101, flow-control module 102 and the He of session management module 103;
Wherein, load balancing control module 101, for the generation in the request selecting flow-control module according to external system Reason server node is issued;Flow-control module 102, including more than one proxy server, for the outside System is attached admissions control and overloading control;Session management module 103, for what is transmitted to the flow-control module Internally server node establishes connection queue and creates session for the request of external system.
Further, the session management module includes more than two session management servers, and the session pipe It is distributed between reason server in a manner of HA;Internal server node more than one is set for bladeization.
The session management module identifies the class of the request message after the request message of the external system is received Type;After the noise in the request message is eliminated according to the type of the request message, taken for request message selection is internal Business device node is issued.
The session management module according to following factor more than one or two establish the external system with it is internal Connection queue between server node:Central processing unit model, network rate, memory rate, application process number, file system Compiler number several, used at the same time.
Further, the session management module calculates the request message number and response message of each internal server node Several differences, balanced highest priority corresponding to the minimum internal server node of the difference, by asking for the external system The internal server node for being sent to balanced highest priority is sought, if more than two inside points server nodes correspond to identical Minimum difference, then selection one is carried out at random from the internal server node with identical minimum difference, as the outside The internal server node of system connection;Or
The session management module is according to the performance label value Y of each internal server noderCarry out the external system Request issues, Yr=Xr/min(X1, X2…Xr…Xy), wherein, XrFor the performance number of default r-th of internal server node, Min represents to take minimum value from the performance number of all internal server nodes, also, the property of each internal server node Energy value is set according to the average duration of each internal server node processing message.
The session that the session management module is additionally operable to establishment is monitored, when monitoring an inside corresponding to session When server node breaks down, by the session equilibrium to other internal server nodes, when monitoring institute corresponding to session When having the internal server node all to break down, the session management module controls the flow-control module to enter offline mould Formula.
Therefore cluster management system provided in an embodiment of the present invention takes the mode of server cluster, by session pipe Reason module issues the control of internal server node for the request message selection of external system, dynamic load balancing is realized, in industry Increase cluster server node when being engaged in occurred frequently and usually reducing clustered node, saved cost of investment, reduced the wasting of resources, Service operation is not interfered with simultaneously.
The embodiment of the present invention also provides a kind of cluster management method, shown in Figure 7 for realizing load balancing, including:
S101, receive the request from external system;
S102, the proxy server nodes issued according to the request selecting of the external system received
S103, the request to the external system are attached admissions control and overloading control;
Internally server node establishes connection queue and creates session for S104, the request to the external system.
It is wherein, above-mentioned that to external system request, internally server node is established connection and included:To the company received Connect request to be authenticated and protocol testing, when the connection authentication and agreement meet preparatory condition, the inside for selecting to issue takes Business device node, the connection request is issued to selected internal server node.
It is described that to outside system request, internally server node is established connection and can also included:Identification is received to ask Seek the type of message;The inside for selecting to issue after the noise in the request message is eliminated according to the type of the request message to take Business device node.Solve the degree of coupling of identification user and request message, the noise portion in balanced message is eliminated by type of message Point, equilibrium treatment is then carried out again.
The index of load balancing includes central processing unit model, network communication situation, memory rate, application process number, text These factors can directly be restrained by part system number and different compilers while service condition etc. by three kinds of strategies:Portfolio Convergence, weighting convergence and time Convergence.
Portfolio restrains:The difference of each node request message and response message is calculated, is worth the balanced preferential height of few person, is worth few person Randomly choosed when not exclusive, if initiation value exception monitoring flow when difference is higher, to solve the problems, such as to apply packet loss difference distortion, Some message is captured during exception monitoring at random and calculates short time consumption to be adjusted;Namely the connection request that the basis receives The internal server node for selecting to issue can also include:The request message number and response for calculating each internal server node disappear The difference of number is ceased, balanced priority corresponding to the few internal server node of the difference is high, if more than two internal sub-services Business device node corresponds to the minimum difference of identical, then is selected at random from the internal server node with identical minimum difference Select.
Weighting convergence:Each node pastes a performance label value, and the high person of performance label value illustrates that node is more reliable, priority Higher, subsequent message amount is in the case of unpredictable, and with label value, high person proceeds by equilibrium.Joint behavior value need to be converted to section Point traffic unit fiducial value, so as to obtain performance label value, conversion formula is Yr=Xr/min(X1, X2…Xr…Xy), wherein, YrPerformance label value is represented, XrFor the performance number of default r-th of internal server node, min is represented from all internal services Minimum value is taken in the performance number of device node, and is converted upwards, conversion upwards is because the high person of value is more reliable.In implementation process, First performance number is sorted, then unit of account fiducial value, namely performance label value, using the high person of performance label value as being compared to each other Benchmark is the step of key one realized, when the least common multiple between there is the relatively low node of performance label value, with performance label value Peak is preferential in node, balanced successively.
Time Convergence:Portfolio Dynamic Programming is realized based on weighting convergence, by the average duration for handling message The performance number of each node is converted to realize.
Further, methods described can also include:Session to establishment is monitored, when monitoring one corresponding to session During individual internal server nodes break down, by the session equilibrium to other internal server nodes, when monitoring session pair When all internal server nodes answered all break down, off-line mode is switched to.
Therefore the cluster management method described in the embodiment of the present invention takes the service request that will receive from server set The node issued is selected in group, is selected according to the performance of request message type and server, realizes dynamic equalization, therefore Increase cluster server node when business is occurred frequently and usually reducing clustered node, saved cost of investment, reduced resource Waste, while do not interfere with service operation.
One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can To instruct the hardware of correlation to complete by program, the program can be stored in a computer-readable recording medium, storage Medium can include:Read-only storage(ROM, Read Only Memory), random access memory(RAM, Random Access Memory), disk or CD etc..
Meanwhile in the above-described embodiments, the description to each embodiment all emphasizes particularly on different fields, and is not described in detail in some embodiment Part, may refer to the associated description of other embodiment.
Cluster management device, management system and the management method provided above the embodiment of the present invention has carried out detailed Jie Continue, specific case used herein is set forth to the principle and embodiment of the present invention, and the explanation of above example is only It is the method and its core concept for being used to help understand the present invention;Meanwhile for those of ordinary skill in the art, according to this hair Bright thought, there will be changes in specific embodiments and applications, in summary, this specification content should not manage Solve as limitation of the present invention.

Claims (10)

  1. A kind of 1. cluster management device, for realizing load balancing, it is characterised in that including:
    Load balancing control module, for the proxy server nodes in the request selecting flow-control module according to external system Issued;
    The flow-control module, including more than one proxy server, for being attached license to the external system Control and overloading control;
    Session management module, including more than two session management servers, it is outer for what is transmitted to the flow-control module Internally server node establishes connection queue and creates session for the request of portion's system;
    Each session management server in the session management module and each proxy server phase in flow-control module Even.
  2. 2. device according to claim 1, it is characterised in that the internal server node is set for bladeization.
  3. 3. device according to claim 1, it is characterised in that the session management module is receiving the external system Request message after, identify the type of the request message;Eliminated according to the type of the request message in the request message Noise;Internal server node is selected to be issued for the request message.
  4. 4. device according to claim 1, it is characterised in that the session management module according to it is following one or two The connection queue that factor above is established between the external system and internal server node:Central processing unit model, network Speed, memory rate, application process number, file system number, compiler number used at the same time.
  5. 5. device according to claim 1, it is characterised in that the session management module passes to the flow-control module Internally server node foundation connection queue includes for the request for the external system sent:The session management module is calculated in each The request message number of portion's server node and the difference of response message number, corresponding to the minimum internal server node of the difference Balanced highest priority, the request of the external system is sent to the internal server node of balanced highest priority, if two Inside points server node more than individual corresponds to the minimum difference of identical, then from the internal server section with identical minimum difference Selection one is carried out in point at random, the internal server node as external system connection;Or
    The session management module is according to the performance label value Y of each internal server noderCarry out the external system request Issue, Yr=Xr/min(X1, X2…Xr…Xy), wherein, XrFor the performance number of default r-th of internal server node, min tables Show and minimum value is taken from the performance number of all internal server nodes, also, the performance number of each internal server node Set according to the average duration of each internal server node processing message.
  6. 6. device according to claim 1, it is characterised in that the session that the session management module is additionally operable to establishment is entered Row monitoring, when monitoring an internal server nodes break down corresponding to session, by the session equilibrium to other interior Portion's server node, when monitoring that all internal server nodes all break down corresponding to session, the session management mould Block controls the flow-control module to enter off-line mode.
  7. A kind of 7. cluster management system, it is characterised in that including:
    More than one external system, asked for requested service processing;
    More than one internal server node, the business processing for handling the external system more than one are asked Ask;
    Cluster management device, the cluster management device include load balancing control module, flow-control module and session management Module;
    Wherein, the load balancing control module, in flow-control module described in the request selecting according to external system Proxy server nodes are issued;Flow-control module, including more than one proxy server, for the outside system System is attached admissions control and overloading control;The session management module, including more than two session management servers, For request internally server node foundation connection queue and the establishment of the external system transmitted to the flow-control module Session;
    Each session management server in the session management module and each proxy server phase in flow-control module Even.
  8. A kind of 8. cluster management method, applied to cluster management device described in claim 1, it is characterised in that including:
    Receive the request from external system;
    The proxy server nodes issued according to the request selecting of the external system received;
    Request to the external system is attached admissions control and overloading control;
    Internally server node establishes connection queue and creates session for request to the external system.
  9. 9. according to the method for claim 8, it is characterised in that the request to the external system internally server Node, which establishes connection queue, to be included:Calculate the request message number of each internal server node and the difference of response message number, institute Balanced highest priority corresponding to stating the minimum internal server node of difference, equilibrium is sent to by the request of the external system The internal server node of highest priority, if more than two inside points server nodes correspond to the minimum difference of identical, Selection one is carried out at random from the internal server node with identical minimum difference as in external system connection Portion's server node;Or
    According to the performance label value Y of each internal server noderCarry out issuing for the external system request, Yr=Xr/min (X1, X2…Xr…Xy), wherein, XrFor the performance number of default r-th of internal server node, min is represented from all internal clothes It is engaged in taking minimum value in the performance number of device node, also, the performance number of each internal server node is according to each internal clothes The average duration for the device node processing message of being engaged in is set.
  10. 10. according to the method for claim 8, it is characterised in that methods described also includes:Session to establishment is supervised Survey, when monitoring an internal server nodes break down corresponding to session, by the session equilibrium to other internal clothes Business device node, when monitoring that all internal server nodes all break down corresponding to session, switches to off-line mode.
CN201310507575.8A 2013-10-24 2013-10-24 A kind of cluster management device, management system and management method Expired - Fee Related CN103685461B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310507575.8A CN103685461B (en) 2013-10-24 2013-10-24 A kind of cluster management device, management system and management method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310507575.8A CN103685461B (en) 2013-10-24 2013-10-24 A kind of cluster management device, management system and management method

Publications (2)

Publication Number Publication Date
CN103685461A CN103685461A (en) 2014-03-26
CN103685461B true CN103685461B (en) 2018-01-30

Family

ID=50321737

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310507575.8A Expired - Fee Related CN103685461B (en) 2013-10-24 2013-10-24 A kind of cluster management device, management system and management method

Country Status (1)

Country Link
CN (1) CN103685461B (en)

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104618466A (en) * 2015-01-20 2015-05-13 上海交通大学 System for balancing load and controlling overload based on message transfer and control method of system
CN105450727B (en) * 2015-11-03 2018-09-18 浪潮(北京)电子信息产业有限公司 A kind of network communication method and network communication architectures
CN105516245B (en) * 2015-11-25 2018-12-14 国家计算机网络与信息安全管理中心 A kind of system and implementation method of the load balancing based on stream
CN105721553B (en) * 2016-01-25 2019-06-25 中国联合网络通信有限公司广东省分公司 A kind of adaptive cluster message distributor
CN106375419A (en) * 2016-08-31 2017-02-01 东软集团股份有限公司 Deployment method and device of distributed cluster
CN106559485B (en) * 2016-11-15 2019-08-06 杭州迪普科技股份有限公司 A kind of method and device of control server shutdown
CN106603443B (en) * 2016-12-14 2019-09-27 郑州云海信息技术有限公司 A method of carrying out data exchange in aggregated server system
CN107704310B (en) * 2017-09-27 2021-06-29 郑州云海信息技术有限公司 Method, device and equipment for realizing container cluster management
CN107979539B (en) * 2017-10-30 2021-12-03 新华三技术有限公司 Message processing method and device
CN108600025B (en) * 2018-05-07 2021-11-02 北京京东尚科信息技术有限公司 Method and device for automatic disaster recovery of system
CN111510958B (en) * 2020-04-27 2024-02-20 中国联合网络通信有限公司广东省分公司 Message access load balancing method and system
CN112738220B (en) * 2020-12-28 2023-02-28 杭州迪普科技股份有限公司 Management method, load balancing method and load balancing equipment of server cluster

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1495634A (en) * 2002-06-27 2004-05-12 上海汉唐科技有限公司 Server clustering load balancing method and system
CN101207561A (en) * 2006-12-22 2008-06-25 华为技术有限公司 Cluster manager, cluster system as well as cluster managing method
CN102387539A (en) * 2011-10-31 2012-03-21 福建三元达通讯股份有限公司 Implement method and system for cluster management and load balancing of WLAN system

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101207550B (en) * 2007-03-16 2010-09-15 中国科学技术大学 Load balancing system and method for multi business to implement load balancing
US20130097322A1 (en) * 2011-10-17 2013-04-18 Alcatel-Lucent Usa, Inc. Scalable distributed multicluster device management server architecture and method of operation thereof

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1495634A (en) * 2002-06-27 2004-05-12 上海汉唐科技有限公司 Server clustering load balancing method and system
CN101207561A (en) * 2006-12-22 2008-06-25 华为技术有限公司 Cluster manager, cluster system as well as cluster managing method
CN102387539A (en) * 2011-10-31 2012-03-21 福建三元达通讯股份有限公司 Implement method and system for cluster management and load balancing of WLAN system

Also Published As

Publication number Publication date
CN103685461A (en) 2014-03-26

Similar Documents

Publication Publication Date Title
CN103685461B (en) A kind of cluster management device, management system and management method
US9948791B2 (en) Sharing group notification
CN103369601B (en) For cell-phone customer terminal provides the method for large concurrent processing and flow control
CN102138307B (en) Method and system for load balancing for services
CN109218355A (en) Load equalizing engine, client, distributed computing system and load-balancing method
WO2014082538A1 (en) Business scheduling method and apparatus and convergence device
WO2015154350A1 (en) Internet access traffic sharing method, device and terminal
CN102893556A (en) Methods, systems, and computer readable media for source peer capacity-based diameter load sharing
CN104811459A (en) Processing method, processing device and system for message services and message service system
CN101459836B (en) Service processing method and system for content distributing network of interactive network television
CN102497379B (en) Network access method, system and equipment
CN105900524A (en) Method and device for managing charging session
CN110266822B (en) Shared load balancing implementation method based on nginx
CN107124453A (en) Platform Interworking GateWay stacks the SiteServer LBS and video call method of deployment
CN108616429A (en) A kind of reconnection method and equipment of Push Service
US11736403B2 (en) Systems and methods for enhanced autonegotiation
CN103905337A (en) Network resource processing device, method and system
CN108234208A (en) The visualization load balancing dispositions method and system of resource management based on business
US20110113131A1 (en) Method of Load Balance Based on BS Architecture
CN101262504A (en) A method, device and system for source and destination IP address translation
CN105721553B (en) A kind of adaptive cluster message distributor
CN111901384B (en) System, method, electronic device and readable storage medium for processing message
CN109600278A (en) A kind of data communications method, device, equipment and medium
CN106411588A (en) Network device management method, master device and management server
CN105188087B (en) A kind of service communication method and system based on Short Message Service Gateway

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
ASS Succession or assignment of patent right

Owner name: HONGKONG SHIYE DEVELOPMENT CO., LTD.

Free format text: FORMER OWNER: CONGXING TECHNOLOGY CO., LTD.

Effective date: 20150729

C41 Transfer of patent application or patent right or utility model
TA01 Transfer of patent application right

Effective date of registration: 20150729

Address after: Room 32, building 3205, Bank of America, 12 Cecil Harcourt Road, central, Hongkong, China

Applicant after: Hongkong world industry development Co., Ltd.

Address before: 24, building 368, 510000 South Guangzhou Avenue, Guangdong, Guangzhou

Applicant before: Sunrise Technology Co., Ltd.

ASS Succession or assignment of patent right

Owner name: TELEFON AB L.M. ERICSSON (SE)

Free format text: FORMER OWNER: HONGKONG SHIYE DEVELOPMENT CO., LTD.

Effective date: 20150908

C41 Transfer of patent application or patent right or utility model
TA01 Transfer of patent application right

Effective date of registration: 20150908

Address after: Stockholm

Applicant after: Telefon AB L.M. Ericsson [SE]

Address before: Room 32, building 3205, Bank of America, 12 Cecil Harcourt Road, central, Hongkong, China

Applicant before: Hongkong world industry development Co., Ltd.

GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20180130

Termination date: 20191024

CF01 Termination of patent right due to non-payment of annual fee