CN104283948A - Server cluster system and load balancing implementation method thereof - Google Patents
Server cluster system and load balancing implementation method thereof Download PDFInfo
- Publication number
- CN104283948A CN104283948A CN201410512754.5A CN201410512754A CN104283948A CN 104283948 A CN104283948 A CN 104283948A CN 201410512754 A CN201410512754 A CN 201410512754A CN 104283948 A CN104283948 A CN 104283948A
- Authority
- CN
- China
- Prior art keywords
- node
- load
- information
- module
- host 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
Classifications
-
- 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
- H04L67/1004—Server selection for load balancing
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Computer And Data Communications (AREA)
Abstract
The invention provides a server cluster system and a load balancing implementation method of the server cluster system. The system comprises a main node, a standby node and common nodes. The main node comprises an information collecting module, a load calculating module and a connecting and distribution deciding module. The standby node is used for reporting the information of the standby node to the main node, and synchronizes with the information of the main node, and the common nodes are screened, so that the selected common node is adopted as a new standby node after the common node becomes the main node. The common nodes are used for reporting the information of the common nodes to the main node, and when one common node becomes the standby node, standby node switching is carried out. Through the server cluster system and the load balancing implementation method, the server cluster system with the main node and the standby node of a mesh network is adopted, the structure of the main node and the standby node is utilized for achieving connection and distribution, the function of sharing connection is achieved, and the problems that connection and distribution delay, and connection is unstable and unreliable can be solved.
Description
Technical field
The present invention relates to server cluster technical field, more specifically, relate to a kind of server cluster system and implementation of load balancing thereof.
Background technology
Along with the frequent generation of the rising of automobile pollution, the aggravation of road congestion conditions and traffic accidents, increasing organization and individual recognize that the driving behavior of the deep driver of understanding will contribute to formulating more rational traffic rule and the more effective intelligent driving navigation system of design, thus reach the object that minimizing traffic accident improves traffic efficiency.And above-mentioned analysis needs to be based upon on the basis of a large amount of travelling datas, need a large amount of intelligent terminal (comprising the equipment such as car-mounted terminal, mobile phone terminal) constantly to gather travelling data in real time and be uploaded to onboard servers and analyze.Therefore, along with the continuous lifting of intelligent terminal and traffic carrying capacity thereof, single server cannot meet the demands, and server cluster becomes the scheme that a kind of needs are dealt with problems.
With regard to intelligent terminal, it carries the important task of Data Collection, and need constantly the travelling data of collection vehicle efficiently, and keep being connected with server, the data collected are sent to service end in real time, and this process is referred to as " data ferry-boat ".In real-time " data ferry-boat " field, the long connected mode of general employing, namely keeps intelligent terminal online for a long time, thus the real-time of guarantee data.Meanwhile, because its data volume depends on the use amount of vehicle, therefore certainly exist with the mass data of server end mutual.Therefore, under this framework, how design server, meets the long connection requirement of a large amount of intelligent terminal on the one hand, ensures the transmitting-receiving of mass data on the other hand and stores normal, becomes the problem that of this field needs to solve.
For the problems referred to above, a solution adopts reverse proxy to realize load balancing.
Adopt reverse proxy realize load balancing scheme in, utilization be Reverse Proxy realize link distribution.Its concrete grammar can be described as: first Reverse Proxy accepts the connection request on automatic network, and connect with it, then the access request from client terminal received subsequently is transmitted to internal server, and the result returned from internal server is transmitted to again corresponding client terminal of initiating connection request.In the process, Reverse Proxy externally shows as a server, is then internally main entrance and the general export of connection request, is responsible for being distributed to each inner server by connecting.
The defect of this solution is: because Reverse Proxy is in the application layer in OSI seven layer model, therefore be necessary for each application service and develop a Reverse Proxy specially, so just limit the range of application of reverse proxy load-balancing technique.At present, reversed proxy server is general all for the load balancing to web server, in addition, in the load balancing that reverse proxy mode realizes, for act of agency each time, proxy server must open two connections, and one externally, and one internally, therefore when concurrent connection request quantity is very large, the load of proxy server is also just very large, in such network, is all the bottleneck that proxy server itself becomes service performance under normal circumstances.
Another kind of the most frequently used solution is the load balancing that address transition Network Based (NAT) technology realizes.At the beginning of NAT technology is born, in order to solve the problem that IPv4 public network address quantity reduces, multiple for inside server is externally mapped as an independent server, makes inner multiple server externally provide service to utilize a public network IP.
In the solution that this address transition Network Based (NAT) technology realizes, from the connection request that network is initiated, carry out address transition through address translation server, then be distributed on each inner server.Adopt the load balancing realized in this way, although reversed proxy server can be solved be positioned at application layer, still need and will develop the problem of Reverse Proxy for often kind of application service specially; And the load balancing that network address translation mode realizes still needs to open two internal and external connections for each request could be realized, time therefore very large in data, address translation server itself can become performance bottleneck.
Meanwhile, in the load balancing that network address translation mode realizes, what realize is certain internal server process request of Dynamic Selection when each request arrives, and needs correctly to find request source when result returns, and therefore needs to retain its conversion mapping table to each request.Same reason, when asking quantitative change large, also needing to consume ample resources to the maintenance of request mapping table, compared with reversed proxy server, more easily causing the performance issue of address translation server; Finally, in the Address Translation services formula load balancing of actual use at present, mostly address translation feature is integrated in hardware switch, its just simple selection strategy realized, the load balancing more optimized and more complicated application protocol can not be supported, and have certain difficulty when expanding.
Summary of the invention
In view of the above problems, the object of this invention is to provide a kind of server cluster system and implementation of load balancing thereof, by using the server cluster system with standby usage node of netted networking, the structure of standby usage node is utilized to realize connecting distribution, reach the function of sharing connection, connect distribution time delay to solve, connect unstable and insecure problem.
On the one hand, the invention provides a kind of server cluster system, comprise host node, secondary node and ordinary node; Wherein,
Host node comprises information collection module, load computing module and is connected distribution decision-making module;
Information collection module, for storing the load information that the secondary node that is connected with host node and each ordinary node report;
Load computing module, for according to the load information reported, obtains the load value of the node corresponding with load information;
Connect distribution decision-making module, for being carried out connection distribution by HTTP redirection, wherein, according to the load value of the node that load computing module obtains, select the node of optimal processing ability, as the node connecting distribution;
Secondary node, for reporting the load information of secondary node to host node, and with the synchronizing information of host node, and screen ordinary node, the ordinary node selected is as new secondary node;
Ordinary node, for reporting the information of described ordinary node to host node, and when it becomes secondary node, is switched to row secondary node.
In addition, preferred scheme is, host node also comprises managing conversation module; For the conversation history record between managing customer terminal and group system, and be responsible for deleting long-time sluggish historical session record between client terminal and group system.In addition, preferred scheme is, creates nodelist, a nodelist for storing the load information of each node be connected with host node in information collection module; Wherein,
Nodelist comprises update mechanism, and update mechanism is for deleting the node of off-line.
In addition, preferred scheme is that load information comprises: the cpu busy percentage of each node be connected with host node, memory usage, network bandwidth occupancy and socket are connected occupancy;
In load computing module, to the load computational algorithm of each node by particular weights, obtain its load capacity that is intensive for CPU intensity, memory-intensive, the network bandwidth and the intensive request of socket respectively, and the result of acquisition is updated to nodelist.
In addition, preferred scheme is, secondary node comprises secondary node information reporting module, synchronizing information module and ordinary node screening module;
Secondary node information reporting module, for reporting load information to host node cycle;
Synchronizing information module, for the load information in periodically synchronizing information collection module, and stores;
Ordinary node screening module, for when secondary node becomes host node, in selection ordinary node, the node of least-loaded becomes that it is alternative.
In addition, preferred scheme is, ordinary node comprises ordinary node information reporting module and secondary node handover module;
Ordinary node information reporting module, for reporting load information to host node cycle;
Secondary node handover module, for become new host node when secondary node after, given notice by the new secondary node gone out selected by it, ordinary node switches to secondary node accordingly, and bears the responsibility of secondary node.
On the other hand, the present invention also provides a kind of implementation of load balancing based on server cluster system, and server cluster system comprises host node, secondary node and ordinary node, wherein:
The load information that host node cycle ground receives coupled secondary node, each ordinary node reports, and according to the load information reported, obtain the load value of each node corresponding with load information;
When there being client terminal connection request, host node carries out connection distribution by HTTP redirection, and wherein, host node, according to the load value of each node obtained, selects the node of optimal processing ability, as the distribution node connected;
When host node breaks down, secondary node carries out active-standby switch, becomes new host node, performs the responsibility of host node, enables new secondary node simultaneously;
Load information is reported in new host node by new secondary node, and with the synchronizing information of new host node, ordinary node is screened simultaneously, selects the alternative of new secondary node.
In addition, preferred scheme is, be provided with session management mechanism in the master node, session management mechanism comprises the conversation history record between client terminal and group system, and be responsible for deleting long-time sluggish historical session record between described client terminal and host node, the node that the connection of selection is distributed is optimal selection.
In addition, preferred scheme is, create a nodelist in the master node, the load information of each node is stored in nodelist, before client's subscriber terminal requests connects,
Judge the node that whether there is off-line in nodelist, if there is the node of off-line, then this node is deleted from nodelist; If the node that there is off-line is secondary node, then again in ordinary node, select new secondary node, and carry out synchronizing information with host node;
If there is not the node of off-line in nodelist, carry out connection distribution according to the connection request of client terminal.
In addition, preferred scheme is, in the process that new secondary node screens ordinary node, new secondary node selects the node of least-loaded alternative as it from ordinary node.
In addition, preferred scheme is that load information comprises: the cpu busy percentage of each node be connected with host node, memory usage, network bandwidth occupancy and socket are connected occupancy;
By the load computational algorithm of particular weights, each node calculates its load capacity that is intensive for CPU intensity, memory-intensive, the network bandwidth and the intensive request of socket respectively, and formula is as follows:
r
l＝f(c
u,m
u,n
u,s
u)
Wherein, r
lrepresent the surplus resources of node;
C
u, m
u, n
u, s
urepresent that cpu busy percentage, memory usage, network bandwidth occupancy and socket that the node received is uploaded connect occupancy respectively.
In addition, preferred scheme is, according to the load information receiving each node and upload, calculate surplus resources percentage, its formula is shown below:
Wherein,
R
lk () represents the surplus resources percentage in k moment;
C
uk () represents k moment CPU usage;
C
u(k-1) utilization rate of (k-1) moment CPU is represented;
At (1-t
u(k)-t
const) in, as t=s, t
uk () represents that the socket in k moment connects utilization rate, t
constrepresent the pre-allowance of socket resource, (1-t
u(k)-t
const) represent the remaining available socket resource situation of node t;
As t=m, t
uk () represents the memory usage in k moment, t
constrepresent the pre-allowance of memory source, (1-t
u(k)-t
const) represent the remaining available memory source situation of node t;
As t=n, t
uk () represents the network bandwidth occupancy in k moment, t
constrepresent the pre-allowance of the network bandwidth, (1-t
u(k)-t
const) represent the remaining available network bandwidth resources situation of node t.
From technical scheme above, server cluster system of the present invention and implementation of load balancing thereof, have following beneficial effect compared with traditional scheme:
First, cluster networking adopts netted networking structure, and possess the progressive relationship of level between multiple node, service stability is improve on the one hand from networking, be beneficial on the other hand and select best performance node in the whole network and become secondary node, thus host node is become under host node failure conditions, ensure global optimum's election;
Secondly, the primary server realized in cluster networking pattern stays field, standby server association control, the layout of common server candidate, active and standbyly directly carries out synchronizing information, under ensureing to reduce synchronous information overhead prerequisite, when host node fault, can realize the seamless switching between host node and secondary node, secondary node can the function of hosting node, shield Single Point of Faliure, improve the availability that cluster externally provides business;
Again, the connection distribution algorithms proposed in host node, under the resource situation considering each node, the node of optimal processing ability is selected to receive service request, the disposal ability of each node can be utilized to greatest extent, it connects distribution mechanisms and meets the demand connecting reasonable distribution on the one hand, greatly reduce the burden of host node server simultaneously, master server is made to carry out active/standby backup co-ordination, for cluster resistance to overturning makes very big contribution, simultaneous session administrative mechanism can reduce time delay when distribution connects further, the reliability of guarantee business and ageing.
In order to realize above-mentioned and relevant object, will describe in detail and the feature particularly pointed out in the claims after one or more aspect of the present invention comprises.Explanation below and accompanying drawing describe some illustrative aspects of the present invention in detail.But what these aspects indicated is only some modes that can use in the various modes of principle of the present invention.In addition, the present invention is intended to comprise all these aspects and their equivalent.
Accompanying drawing explanation
By reference to the content below in conjunction with the description of the drawings and claims, and understand more comprehensively along with to of the present invention, other object of the present invention and result will be understood and easy to understand more.In the accompanying drawings:
Fig. 1 is the load balancing networking structure schematic diagram that reverse proxy and network address transmission technology realize;
Fig. 2 is the server cluster system networking structure schematic diagram according to the embodiment of the present invention;
Fig. 3 is the server cluster system logical construction block diagram according to the embodiment of the present invention;
Fig. 4 is the host node logical construction block diagram according to the embodiment of the present invention;
Fig. 5 is the server cluster system Application Example schematic diagram according to the embodiment of the present invention;
Fig. 6 is the implementation of load balancing schematic flow sheet based on server cluster system according to the embodiment of the present invention;
Fig. 7 is the principle process schematic diagram of client terminal according to the embodiment of the present invention and group system connection request process;
Fig. 8 is the workflow schematic diagram realized according to the load balancing of the server cluster system of the embodiment of the present invention.
Label identical in all of the figs indicates similar or corresponding feature or function.
Embodiment
In the following description, for purposes of illustration, in order to provide the complete understanding to one or more embodiment, many details have been set forth.But, clearly, also these embodiments can be realized when there is no these details.
In the load balancing networking structure that reverse proxy and network address transmission technology realize, its network construction form is all the trees adopting single-point entry, Fig. 1 shows the load balancing networking structure of reverse proxy and the realization of network address transmission technology, as shown in Figure 1, by the main entrance of a server as flow, other inner servers are given by traffic distribution, then by the pass traffic returned to connecting originating end.Meanwhile, the task born due to portal server is more, and general being difficult to carries out active/standby backup to it, and the active-standby switch time is also difficult to meet the demands, and wherein, active/standby backup is exactly that secondary node periodically backs up the load information that host node receives.
In Address Translation services formula load balancing for actual use at present, only has simple selection strategy, the problem of load balancing and the more complicated application protocol more optimized can not be supported, the present invention by analysis reversed proxy server realize load balancing, the schemes such as the load balancing that network address transmission technology realizes, on the basis analyzing all kinds of scheme pluses and minuses, in conjunction with business scenario, propose a kind of new server cluster system and implementation of load balancing, what server cluster system provided by the invention adopted is netted networking structure, and on the basis carefully analyzing used agreement, utilize the specific function of agreement to achieve dexterously innovatively and connect distribution, reach the function of sharing connection, compared with connecting distribution algorithms with tradition, there is time delay little, the feature of stable connection, adapt to the characteristic of vehicle-mounted client terminal fast moving.
Below with reference to accompanying drawing, specific embodiments of the invention are described in detail.
In order to the network structure of server cluster system provided by the invention is described, Fig. 2 shows the netted networking structure of the server cluster system according to the embodiment of the present invention.
As shown in Figure 2, the networking structure of server cluster system that the present invention proposes is netted networking structure, wherein, establishes three kinds of node types, host node, secondary node and ordinary node in the cluster service framework possessing active and standby function.
Cluster networking of the present invention improves traditional group net mode from two aspects, first, alleviate the workload burden of portal server, make it can meet the requirement of master and backup network, realize primary server and stay field, standby server association control, the layout of common server candidate, shield Single Point of Faliure, drastically increase networking stability; Secondly, tree-like networking structure is replaced with netted networking structure, improves service stability on the one hand from networking, being beneficial to best performance node in election the whole network on the other hand becomes secondary node, thus host node is become under host node failure conditions, ensure global optimum's election.
The netted group-network construction of cluster used in the present invention is described in detail in detail below.
As can be seen from Figure 2, the task that the scheme proposing secondary node is mainly born based on the portal server of current load balancing is more, general being difficult to carries out active/standby backup to it, the active-standby switch time is also difficult to meet the demands, and therefore under this mode, proxy server will inevitably become the important node of Single Point of Faliure; And active and standby in the present invention between synchronizing information be exactly can rapid pipe connecting work for the ease of secondary node during host node fault, if only there is host node and ordinary node in net, want to realize seamless switching, principal point carries out synchronizing information with regard to ordinary node in needs and net, and the expense of in this case synchronous amount of information can than only large in secondary node synchronization overhead.Below each node is briefly introduced, to understand overall networking mode.
Host node is responsible for the nodal information collecting each node in cluster, and calculates through overload, calculates the load value of each node, the connection request of host node managing customer terminal, connects carry out connection distribution according to certain rule to a large amount of client terminals; Meanwhile, host node sets up session management mechanism, sets up the express passway between functional client terminal and server cluster system, lowers time delay further.
Secondary node mainly realizes three functions, one is that the load information of oneself node is periodically reported host node, the foundation connecting distribution node is selected as host node, secondary node is periodically from host node synchronized clusters internal burden information simultaneously, combine with host node and realize active/standby backup, under host node abnormal conditions, secondary node switches to host node, avoids causing group system externally can not provide the situation of business because host node is abnormal; Finally, secondary node is responsible for best performance node in election ordinary node, to become secondary node after oneself becomes host node.
Ordinary node realizes load information cycle reporting mechanism, is responsible for reporting loading condition to host node cycle, and meanwhile, its most important task is the link processing the client terminal be redirected by host node, for client terminal provides service.
Described by the composition function of the server cluster system in Fig. 2, server cluster system provided by the invention, host node is responsible for the professional chief of smart client terminal to connect balancedly to break up in cluster on each node, avoid the performance bottleneck that single node brings, secondly, session management mechanism on host node can realize connecting and redirect fast, reduces time delay; The back mechanism of standby usage node associating, when can realize host node fault, secondary node can become new host node fast, group system break down node time still can be stable business service is provided.After secondary node becomes host node, new secondary node will be selected, and whole group system becomes stratified system of going forward one by one, as long as there is a station server normally to work, outwards can provide normal service, maximize the availability improving cluster.
According to netted networking structure provided by the invention, the present invention classifies to the node in server cluster system, always be divided into host node, secondary node and ordinary node three class, each category node collaborate, the division of labor is clear and definite, while externally providing service, achieve cluster internal load balancing, improve entire system stability.Fig. 3 shows the server cluster system logical construction according to the embodiment of the present invention.
As shown in Figure 3, the invention provides a kind of server cluster system 300, comprising: host node 310, secondary node 320 and ordinary node 330.
Wherein, host node 310 comprises information collection module 311, load computing module 312 and is connected distribution decision-making module 313.
Secondary node 320 for reporting the load information of secondary node to host node, and with the synchronizing information of host node, and screen ordinary node, the ordinary node selected is as new secondary node.
Ordinary node 330 for reporting the load information of ordinary node to host node, and when it becomes secondary node, is switched to secondary node.
In order to describe the logical construction of the host node in server cluster system in detail, Fig. 4 shows the host node logical construction according to the embodiment of the present invention, as shown in Figure 4, host node 310 comprises information collection module 311, load computing module 312 further, connects distribution decision-making module 313 and session management module 314.
Particularly, the load information that reports for storing the secondary node that is connected with host node and ordinary node of information collection module 311.
Load computing module 312, for according to the load information reported, obtains the load value of the node corresponding with load information.
Connect distribution decision-making module 313 for being carried out connection distribution by HTTP redirection, wherein, according to the load value of the node that load computing module obtains, select the node of optimal processing ability, as the node connecting distribution.
Managing conversation module 314 for the conversation history record between managing customer terminal and group system, and is responsible for deleting long-time sluggish historical session record between client terminal and group system.
Wherein, in information collection module 311, create a nodelist (node listing), the load information that nodelist reports for storing each node of being connected with host node; Wherein, nodelist comprises update mechanism, and update mechanism is for deleting the node of off-line.
Load information comprises: (socket is the intervening software abstraction layer that application layer communicates with TCP/IP protocol suite for the cpu busy percentage of each node be connected with host node, memory usage, network bandwidth occupancy and socket, it is a group interface, concerning user, one group of simple interface is exactly whole, socket is allowed to remove organising data, to meet the agreement of specifying) connect occupancy; In load computing module 312, to the load computational algorithm of each node by particular weights, obtain its load capacity that is intensive for CPU intensity, memory-intensive, the network bandwidth and the intensive request of socket respectively, and the result of acquisition is updated to nodelist.
Secondary node 320 comprises secondary node information reporting module, synchronizing information module and ordinary node screening module further.
Particularly, secondary node information reporting module, for reporting load information to host node cycle.
Synchronizing information module, for the load information in periodically synchronizing information collection module, and stores.
Ordinary node screening module, for when secondary node becomes host node, in selection ordinary node, the node of least-loaded becomes that it is alternative.
Ordinary node comprises ordinary node information reporting module and secondary node handover module.
Particularly, ordinary node information reporting module, for reporting load information to host node cycle.
Secondary node handover module, for become new host node when secondary node after, given notice by the new secondary node gone out selected by it, ordinary node switches to secondary node accordingly, and bears the responsibility of secondary node.
In addition, in the present invention, be described in detail for the server cluster system of five node compositions, Fig. 5 shows the server cluster system Application Example structure according to the embodiment of the present invention.
As shown in Figure 5, A is host node, and B is secondary node, and C, D, E are ordinary node.Wherein, host node A comprises information collection module, load computing module, connects distribution decision-making module and session management module.Secondary node B comprises secondary node information reporting module, synchronizing information module and ordinary node screening module.Ordinary node C, ordinary node D, ordinary node E comprise ordinary node information reporting module and secondary node handover module respectively.
In the information collection module of host node A as shown in Figure 5, host node A is responsible for and secondary node B and keep stable long connection status between ordinary node C, D, E, obtains the state information of each node with this.A node creates a nodelist, so that the load information that each node that storage is attached thereto reports, when A node receives the state information that B, C, D, E periodically report, (state information comprises: the load information of each node, namely cpu busy percentage, memory usage, network bandwidth occupancy and socket connect occupancy), corresponding load information is updated in the nodelist that this node preserves by A node, and using these information as the selection gist connecting distribution and sentence timed node.Nodelist comprises update mechanism, and update mechanism ensure that deletes the ossified node of not online (off-line) in good time.As: when host node A does not receive the corresponding information of a node for a long time, namely think this one malfunctions, therefore the information of this node is rejected from nodelist, when avoiding A node to carry out connection distribution, say that connection is distributed on this node.
In the load computing module of host node A as shown in Figure 5, the node load information that A node was collected at information collection module, comprise CpuUseRate, MemUseRate, NetworkUseRate, SocketUseRate, represent the cpu busy percentage of each node, memory usage, network bandwidth occupancy and socket respectively and connect occupancy.A node calculates according to the load value of the load information collected to each node, loading algorithm will adopt the load computational algorithm of particular weights, that is: for each node, calculate its load capacity that is intensive for CPU intensity, memory-intensive, the network bandwidth and the intensive request of socket respectively, effectively to process different request types, more reasonably carry out connection distribution, the result calculated is updated in the nodelist of corresponding node and goes, be convenient to follow-up carry out rapidly connection distribution.
In the connection distribution decision-making module of host node A as shown in Figure 5, when there being intelligent terminal connection request, select the node of optimal processing ability, and connection is distributed on this node, assuming that in B, C, D, E tetra-nodes, select the node calculating optimal processing ability in load computing module to be D node, then the connection of intelligent terminal is redirected on D node by A node, by this scheme, can will connect balanced breaing up on each node, and the time delay of process can be reduced.The detailed process and the related algorithm that connect distribution will be further described below.
In the session management module of host node A as shown in Figure 5, host node sets up session management mechanism, the conversation history record of management intelligent terminal and server cluster system, distributes fast for the follow-up connection of intelligent terminal, reduces request time delay to greatest extent.Session management module is responsible for deleting those long-time sluggish conversation recordings, thus ensures that bringing back to life connection can break up according to connection distribution algorithms, makes connection distribute more reasonable.
That is, session management module will delete those long-time sluggish conversation recordings, be due to intelligent terminal be connected the node distributed and do not connect for a long time, the load of the node in group system can change, the load of the node of connection distribution originally also can change, therefore can not ensure that the node of original connection distribution is the node of optimal processing ability, therefore this conversation recording is deleted, reselect the node of new connection distribution, with the node ensureing that the node selected is optimal processing ability.Therefore session management module reduces time delay when distribution connects further, ensures the reliability of business between intelligent terminal and group system and ageing.
In the secondary node information reporting module of secondary node B as shown in Figure 5, secondary node B periodically reports oneself load information to host node A, its load information comprises cpu load, internal memory load, network bandwidth load and socket load etc., facilitates host node A to utilize these information to carry out connection distribution and node administration.
In the synchronizing information module of secondary node B as shown in Figure 5, secondary node B is synchronous nodelist information from host node A periodically, is kept in the internal memory of this node, the information to group system between synchronous standby usage node.Secondary node B adopts timing mechanism and event trigger mechanism to detect the presence of host node A: first, and timing is detected the presence of host node A by host node A; Secondly, when periodically reporting information is to host node A failure for secondary node B, triggering is detected the presence of host node A, if testing result is host node A off-line, then secondary node B will switch to host node, and each node in mass-sending notice group system, notify that the ordinary node selected in advance becomes secondary node, thus continue to provide service to the connection request of intelligent terminal.
In the ordinary node screening module of secondary node B as shown in Figure 5, secondary node B becomes the alternative of oneself by selecting the node of least-loaded with longer cycle, and when oneself becoming host node, notify that this node becomes secondary node, thus the repairing of the networking structure of predetermined network is completed when host node lost efficacy, continue the stability ensureing network, externally provide service.
In the ordinary node information reporting module of ordinary node as shown in Figure 5, ordinary node C, D, E are periodically to the load information of host node report oneself, comprise cpu load, internal memory load, network bandwidth load and socket load etc., facilitate host node to utilize these information to carry out connection distribution and node administration.
In the secondary node handover module of ordinary node as shown in Figure 5, when after the host node that secondary node B becomes new, the new secondary node gone out selected by it is given notice, ordinary node C carries out secondary node switching accordingly, comprise the various functions and service of bearing secondary node, complete current connection request and return results, communicating with host node subsequently, continuing to manage the node in cluster.
Above-mentioned is the function of each node in server cluster system, and in order to the realization of the load-balancing function of each node in server cluster system is described, Fig. 6 shows the implementation of load balancing flow process based on server cluster system according to the embodiment of the present invention.
As shown in Figure 6, the method that the load balancing based on server cluster system provided by the invention realizes, server cluster system comprises host node, secondary node and ordinary node.Wherein,
S610: the load information that host node cycle ground receives coupled secondary node, each ordinary node reports.
Wherein, it should be noted that, create a nodelist in the master node, the load information of each node is stored in nodelist, and load information comprises: the cpu busy percentage of each node be connected with host node, memory usage, network bandwidth occupancy and socket are connected occupancy.
S620: host node, according to the load information reported, obtains the load value of each node corresponding with load information.
Before subscriber terminal requests connects, judge the node that whether there is off-line in nodelist, if there is the node of off-line, then this node is deleted from nodelist; If the node that there is off-line is secondary node, then again in ordinary node, select new secondary node, and carry out synchronizing information with host node; If there is not the node of off-line in nodelist, carry out connection distribution according to the connection request of described client terminal.
S630: when there being client terminal connection request, host node carries out connection distribution by HTTP redirection, and wherein, host node, according to the load value of each node obtained, selects the node of optimal processing ability, as the distribution node connected.
Wherein, be provided with session management mechanism in the master node, session management mechanism comprises the conversation history record between client terminal and group system, and is responsible for deleting long-time sluggish historical session record between client terminal and host node, and the node that the connection of selection is distributed is optimal selection.
This be due to client terminal be connected the node distributed and do not carry out business contact for a long time, the load of each node in group system can change, for ensureing that the node of service connection distribution is the node of optimal processing ability, therefore conversation recording is deleted, reselect the node of new connection distribution.Session management mechanism the time delay reduced further when distribution connects is set, ensure the reliability of business between intelligent terminal and group system and ageing.
S640: when host node breaks down, secondary node carries out active-standby switch, becomes new host node, performs the responsibility of host node, enables new secondary node simultaneously.
S650: load information is reported in new host node by new secondary node, and with the synchronizing information of new host node, ordinary node is screened simultaneously, selects the alternative of new secondary node.
Above-mentioned steps S640 and step S650 is the workflow of active-standby switch.When host node breaks down, secondary node how to replace host node to carry out fulfiling host node responsibility.The back mechanism of host node and secondary node associating, when can realize host node fault, secondary node can become new host node fast, still stably can provide business service when server cluster system goes wrong.
Detailed description host node is gone out the detailed process of node as the distribution node connected of optimal processing ability below.
The present invention adopts and backs up host node in server cluster system, have the following advantages compared with the load balancing that this back mechanism and reversed proxy server and network address change technology realize: on the one hand, the host node in server cluster system does not need to bear application layer work as Reverse Proxy; On the other hand, do not need to set up as network address transmission to be bi-directionally connected yet, because this reducing the workload of host node, reducing individual node failure risk, making active and standby realization become possibility yet, improving cluster stability further.Operation principle and respective algorithms that portal server (i.e. host node) connects distribution will be described below.
As everyone knows, in http protocol, HTTP request is processed by service end after arriving service end, and returns HTTP response, and this response can carry HTTP return code, and each return code has contained different implications, as shown in table 1.
Table 1
| HTTP return code | Implication |
| 1** | Retain |
| 2** | Represent that request successfully receives |
| 3** | Be redirected |
| 4** | Client terminal mistake |
| 5** | Service end mistake |
Wherein, show that network site has been changed in asked service with the HTTP return code of 3 beginnings, need client terminal end again to initiate request, new chained address will return in the lump in HTTP responds.
Host node utilizes in HTTP return code 307, and its implication for adopting interim being redirected, shows the temporarily better position of asked service.The interim benefit be redirected is used to have: the network address that can return hope in HTTP responds, allows client terminal initiate connection request to new address then; Secondly, owing to being interim being redirected, client terminal is when initiating new connection request, can not to be redirected after address preserve, so that during each connection request, first host node all can receive the connection request of client terminal, then carries out connection distribution according to the loading condition of each node, ensure that the realization of clustering design function.
Fig. 7 shows the principle process of client terminal according to the embodiment of the present invention and group system connection request process.In the present embodiment, using concrete vehicle-mounted client as client terminal.
As shown in Figure 7, vehicle-mounted client sends connection request to host node, host node obtains the ordinary node C of the connection distribution of optimal processing ability by connecting distribution algorithms, the information of the ordinary node C of optimal processing ability that vehicle-mounted client will return according to host node, set up professional chief to ordinary node C to connect, then vehicle-mounted client and ordinary node C start Business Processing.
In the present invention, host node carries out connection distribution by HTTP redirection, undertaken connecting distribution by HTTP redirection and have following benefit: host node only needs to carry out the process of a HTTP request, do not provide service to application, also internally do not connect, dramatically saves on the process resource of host node, alleviate the burden of host node; Simultaneously, the new address of carrying due to HTTP redirection is the address by the internal node calculated, there is unpredictability, the identifying code that its functional equivalent uses when access to netwoks, the ineffective access initiated by automaton can be shielded to a certain extent, therefore there is the ability of anti-DDoS network attack, improve the fail safe of entire system.
In connection distribution algorithms, the most important thing (that is: selecting the node of optimal processing ability) to the selection of the node of least-loaded, can it not only relate to connection request and be processed fast, also relate to the stability of group system entirety, will be described least-loaded node selection algorithm below.
Due to a nodelist will be created on the primary node, real-time update every loading condition of each node, the data four-tuple that each node was collected, be specially CPUUsage, MemUsage, NetworkUsage and SocketUsage, represent the consumption level of CPU, internal memory, the network bandwidth and Socket number of connection on node respectively.According to the difference of connection request, connection type is divided into CPU intensity, memory-intensive and bandwidth intensive.Therefore, by the load computational algorithm of particular weights, each node will calculate its disposal ability to different connection request respectively, as follows:
r
l＝f(c
u,m
u,n
u,s
u)
Wherein, r
lrepresent the surplus resources of node; When remaining cpu resource and being larger, show that the workload that it can be born is more, then can be distributed to more connecting respective nodes and get on.C
u, m
u, n
u, s
urepresent that cpu busy percentage, memory usage, network bandwidth occupancy and socket that the node received is uploaded connect occupancy respectively, be worth larger, show that corresponding resource occupation is larger, node load is heavier.According to the corresponding data received, calculate final surplus resources percentage, its formula is shown below:
Wherein, r
lk () represents the surplus resources percentage in k moment; c
uk () represents k moment CPU usage; c
u(k-1) utilization rate of (k-1) moment CPU is represented; The CPU impact factor considers two moment, utilizes the method for discreet derivative to estimate the change curve of CPU.
At (1-t
u(k)-t
const) in, as t=s, t
uk () represents that the socket in k moment connects utilization rate, t
constrepresent the pre-allowance of socket resource, (1-t
u(k)-t
const) represent the remaining available socket resource situation of node t; As t=m, t
uk () represents the memory usage in k moment, t
constrepresent the pre-allowance of memory source, (1-t
u(k)-t
const) represent the remaining available memory source situation of node t; As t=n, t
uk () represents the network bandwidth occupancy in k moment, t
constrepresent the pre-allowance of the network bandwidth, (1-t
u(k)-t
const) represent the remaining available network bandwidth resources situation of node t.
It should be noted that, generally, the pre-allowance of socket resource and network broadband resource is all 0, show that the utilization rate of this resource can arrive 100%, and it is corresponding, the pre-allowance of memory source can not be 0, shows that the utilization rate of memory source can not arrive 100%, needs to retain certain space.
That is, the surplus resources percentage r in k moment
lk () represents, it is subject to c respectively
u, m
u, n
u,
the impact of individual factor of influence, the influence mode of each factor of influence to it is different; And the influence mode of each factor of influence to it is different, if one of them factor of influence reaches capacity, if cpu surplus is 0, then the resource end value calculated will become 0, then this node can not obtain Resourse Distribute, ensure that the stability of this node and the correct response of access request.In 4 factors of influence, according to the different qualities of every class resource, the influence mode of its correspondence is also different.
For socket connection, 1-s
u(k)-s
constrepresent the factor of influence that socket connects; s
uk socket that () is the k moment connects occupancy; s
constfor the pre-allowance of socket resource.Generally, the pre-allowance of socket resource and network resource is all 0, show that the utilization rate of these resources can arrive 100%, and it is corresponding, the pre-allowance of Mem resource can not be 0, shows that the utilization rate of memory source can not arrive 100%, and this is the normal operation in order to ensure node, all need for internal memory retains certain space, in order to avoid cause node to crash.
Wherein, it should be noted that, in the calculating of factor of influence, the calculating of more special is cpu resource, from system level, cpu resource is very valuable resource, be again the resource of utilization rate moment change simultaneously, cpu resource can not be wasted in the mode of simple headspace, the curve that improves of CPU must be considered again, also need the complexity considering to calculate, to save the computational resource of host node in periodically a large amount of calculating simultaneously.
In sum, utilize the method for discreet derivative to estimate the change curve of CPU, and it can be used as the headspace of CPU to consider, when CPU curve presents ascendant trend, its first derivative is greater than 0, it can be used as CPU to use pre-allowance to use, draws 1-2c
u(k)+c
u(k-1) factor of influence computing formula, on the contrary when CPU usage curve presents downward trend, its first derivative is negative value, because CPU will obtain more surplus, is not now its reserved utilization rate space.Adopt the CPU use amount of curvilinear path Forecasting Methodology estimation, can not only meet the estimation requirement of CPU usage, at utmost can improve the utilization rate of CPU in Shangdi, elevator system overall efficiency, have received good effect simultaneously.
In order to further illustrate server cluster system node between how to realize load balancing, Fig. 8 shows the workflow of the implementation of load balancing of the server cluster system according to the embodiment of the present invention.In the present embodiment, using concrete vehicle-mounted client as client terminal.
As shown in Figure 8, the implementation of load balancing of the server cluster system shown in the present embodiment, comprises the steps:
S801: start;
S802: host node exists fault? if there is fault, perform step S803; If there is not fault, perform step S808;
S803: active-standby switch, secondary node becomes new host node;
S804: new secondary node is enabled;
S805: secondary node information reporting;
S806: with host node synchronizing information;
S807: secondary node carries out ordinary node screening; After active-standby switch completes, fulfil the responsibility of host node; Perform step S808 again;
S808: host node cycle receives other node load information reporting;
S809: host node calculates the load value of each node according to the loading condition reported;
Not line node is there in S810:nodelist? if there is not online node, perform step S711; If there is no not online node, perform step S815;
S811: not online node is secondary node? if secondary node, perform step S712; If not, perform step S814;
S812: remove this node from nodelist;
S813: reselect new secondary node and carry out synchronizing information, then performing step S815;
S814: remove this node from nodelist, then perform step S815;
S815: when there being vehicle-mounted client's side link request, selects the node of optimal processing ability as the node connecting distribution;
S816: the node of the connection distribution that vehicle-mounted client and host node are selected starts to carry out operation flow;
S817: terminate.
Above-mentioned is the flow process of a specific embodiment of the implementation of load balancing of server cluster system.In sum, the present invention is directed to the concurrent long feature connected of height of intelligent terminal, propose this long connection, high concurrent services end implementation of load balancing and server cluster system and connect distribution algorithms, server cluster system of the present invention adopts netted networking structure, and possess the progressive relationship of level between multiple node, service stability is improve on the one hand from networking, be beneficial to best performance node in election the whole network on the other hand and become secondary node, thus host node is become under host node failure conditions, ensure global optimum's election.
Further, in cluster networking pattern, adopt active/standby backup mechanism, can realize active and standby between seamless switching, secondary node can the function of hosting node, shields Single Point of Faliure, improves the availability that cluster externally provides business.
Again, the connection distribution algorithms proposed, selects the node of optimal processing ability to receive service request, can utilize the disposal ability of each node to greatest extent, simultaneous session administrative mechanism can reduce time delay when distribution connects further, ensures the reliability of business and ageing.Therefore, the method that the present invention proposes load balancing realization has the possibility extensively used, and will play even more important effect in all kinds of similar project.
The server cluster system and implementation of load balancing thereof that propose according to the present invention is described in an illustrative manner above with reference to accompanying drawing.But, it will be appreciated by those skilled in the art that server cluster system and implementation of load balancing thereof that the invention described above is proposed, various improvement can also be made on the basis not departing from content of the present invention.Therefore, protection scope of the present invention should be determined by the content of appending claims.
Claims (12)
1. a server cluster system, comprises host node, secondary node and ordinary node; Wherein,
Described host node comprises information collection module, load computing module and is connected distribution decision-making module;
Described information collection module, for storing the load information that the secondary node that is connected with described host node and each ordinary node report;
Described load computing module, for the load information reported described in basis, obtains the load value of the node corresponding with described load information;
Described connection distribution decision-making module, for being carried out connection distribution by HTTP redirection, wherein, according to the load value of the node that described load computing module obtains, selects the node of optimal processing ability, as the node connecting distribution;
Described secondary node, for reporting the load information of secondary node to described host node, and with the synchronizing information of described host node, and screen described ordinary node, the ordinary node selected is as new secondary node;
Described ordinary node, for reporting the load information of ordinary node to described host node, and when it becomes secondary node, is switched to secondary node.
2. server cluster system as claimed in claim 1, wherein,
Described host node also comprises managing conversation module, for the conversation history record between managing customer terminal and described group system, and is responsible for deleting long-time sluggish historical session record between described client terminal and described group system.
3. server cluster system as claimed in claim 1, wherein,
The load information that a nodelist, described nodelist report for storing each node of being connected with described host node is created in described information collection module; Wherein,
Described nodelist comprises update mechanism, and described update mechanism is for deleting the node of off-line.
4. server cluster system as claimed in claim 3, wherein,
Load information comprises: the cpu busy percentage of each node be connected with described host node, memory usage, network bandwidth occupancy and socket are connected occupancy;
In described load computing module, to the load computational algorithm of each node by particular weights, obtain its load capacity that is intensive for CPU intensity, memory-intensive, the network bandwidth and the intensive request of socket respectively, and the result of acquisition is updated to described nodelist.
5. server cluster system as claimed in claim 1, wherein,
Described secondary node comprises secondary node information reporting module, synchronizing information module and ordinary node screening module;
Described secondary node information reporting module, for reporting load information to described host node cycle;
Described synchronizing information module, for the load information in periodically synchronous described information collection module, and stores;
Described ordinary node screening module, for when described secondary node becomes host node, in selection ordinary node, the node of least-loaded becomes that it is alternative.
6. server cluster system as claimed in claim 1, wherein,
Described ordinary node comprises ordinary node information reporting module and secondary node handover module;
Described ordinary node information reporting module, for reporting load information to described host node cycle;
Described secondary node handover module, for when after the host node that described secondary node becomes new, given notice by the new secondary node gone out selected by it, ordinary node switches to secondary node accordingly, and bears the responsibility of secondary node.
7., based on an implementation of load balancing for server cluster system, described server cluster system comprises host node, secondary node and ordinary node, wherein,
The described load information that host node cycle ground receives coupled secondary node, each ordinary node reports, and according to the described load information reported, obtain the load value of each node corresponding with described load information;
When there being client terminal connection request, described host node carries out connection distribution by HTTP redirection, and wherein, described host node, according to the load value of each node obtained, selects the node of optimal processing ability as the node connecting distribution;
When described host node breaks down, described secondary node carries out active-standby switch, becomes new host node, performs the responsibility of host node, enables new secondary node simultaneously;
Load information is reported in described new host node by described new secondary node, and with the synchronizing information of new host node, described ordinary node is screened simultaneously, selects the alternative of described new secondary node.
8., as claimed in claim 7 based on the implementation of load balancing of server cluster system, wherein, in described host node, be provided with session management mechanism,
Described session management mechanism comprises the conversation history record between described client terminal and described group system, and be responsible for deleting long-time sluggish historical session record between described client terminal and described host node, the node that the connection of selection is distributed is optimal selection.
9. as claimed in claim 7 based on the implementation of load balancing of server cluster system, wherein, in described host node, create a nodelist, the load information of each node is stored in described nodelist, before described subscriber terminal requests connects,
Judge the node that whether there is off-line in described nodelist, if there is the node of off-line, then this node is deleted from described nodelist; If the node that there is off-line is secondary node, then again in described ordinary node, select new secondary node, and carry out synchronizing information with described host node;
If there is not the node of off-line in described nodelist, carry out connection distribution according to the connection request of described client terminal.
10. as claimed in claim 7 based on the implementation of load balancing of server cluster system, wherein, in the process that described new secondary node screens described ordinary node,
Described new secondary node selects the node of least-loaded alternative as it from described ordinary node.
11. as claimed in claim 7 based on the implementation of load balancing of server cluster system, wherein,
Described load information comprises: the cpu busy percentage of each node be connected with described host node, memory usage, network bandwidth occupancy and socket are connected occupancy;
By the load computational algorithm of particular weights, each node calculates its load capacity that is intensive for CPU intensity, memory-intensive, the network bandwidth and the intensive request of socket respectively, and formula is as follows:
r
l＝f(c
u,m
u,n
u,s
u)
Wherein, r
lrepresent the surplus resources of node;
C
u, m
u, n
u, s
urepresent that cpu busy percentage, memory usage, network bandwidth occupancy and socket that the node received is uploaded connect occupancy respectively.
12. as claimed in claim 11 based on the implementation of load balancing of server cluster system, wherein,
According to the load information receiving each node and upload, calculate surplus resources percentage, its formula is shown below:
Wherein, r
lk () represents the surplus resources percentage in k moment;
C
uk () represents k moment CPU usage;
C
u(k-1) utilization rate of (k-1) moment CPU is represented;
At (1-t
u(k)-t
const) in, as t=s, t
uk () represents that the socket in k moment connects utilization rate, t
constrepresent the pre-allowance of socket resource, (1-t
u(k)-t
const) represent the remaining available socket resource situation of node t;
As t=m, t
uk () represents the memory usage in k moment, t
const represents the pre-allowance of memory source, (1-t
u(k)-t
const) represent the remaining available memory source situation of node t;
As t=n, t
uk () represents the network bandwidth occupancy in k moment, t
constrepresent the pre-allowance of the network bandwidth, (1-t
u(k)-t
const) represent the remaining available network bandwidth resources situation of node t.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410512754.5A CN104283948B (en) | 2014-09-26 | 2014-09-26 | Server cluster system and its implementation of load balancing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410512754.5A CN104283948B (en) | 2014-09-26 | 2014-09-26 | Server cluster system and its implementation of load balancing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104283948A true CN104283948A (en) | 2015-01-14 |
| CN104283948B CN104283948B (en) | 2018-12-07 |
Family
ID=52258421
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410512754.5A Active CN104283948B (en) | 2014-09-26 | 2014-09-26 | Server cluster system and its implementation of load balancing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104283948B (en) |
Cited By (47)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104768196A (en) * | 2015-04-21 | 2015-07-08 | 四川西结微波科技发展有限责任公司 | Cluster group seamless switching method |
| CN105007233A (en) * | 2015-07-13 | 2015-10-28 | 互联网域名系统北京市工程研究中心有限公司 | Method for distributing address based on DHCP (dynamic host configuration protocol) server cluster load |
| CN105338078A (en) * | 2015-10-26 | 2016-02-17 | 北京百度网讯科技有限公司 | Data storage method and device used for storing system |
| CN105430105A (en) * | 2016-01-05 | 2016-03-23 | 北京卓识达软件有限公司 | Data processing method and system for vehicle-mounted terminal |
| CN105516245A (en) * | 2015-11-25 | 2016-04-20 | 国家计算机网络与信息安全管理中心 | Flow-based load balancing system and realization method thereof |
| CN105577759A (en) * | 2015-12-15 | 2016-05-11 | 东软熙康健康科技有限公司 | Server node allocation method and device |
| CN105989279A (en) * | 2015-02-13 | 2016-10-05 | 上海通用识别技术研究所 | Domino dedicated password cracking system based on multi-hierarchy GPU (Graphic Process Unit) cluster |
| CN106027649A (en) * | 2016-05-20 | 2016-10-12 | 深圳市永兴元科技有限公司 | Data collection method of distributed data system and the distributed data system |
| CN106060123A (en) * | 2016-05-20 | 2016-10-26 | 深圳市永兴元科技有限公司 | Distributed data system data acquisition method and distributed data system |
| CN106331755A (en) * | 2016-08-17 | 2017-01-11 | 广州酷狗计算机科技有限公司 | Method and device for processing media stream blockage |
| CN106453120A (en) * | 2015-08-05 | 2017-02-22 | 北京网御星云信息技术有限公司 | Dynamic clustering method and dynamic clustering system |
| CN106549796A (en) * | 2016-09-27 | 2017-03-29 | 努比亚技术有限公司 | Resource control method and host node that a kind of firmware space is downloaded |
| CN106716974A (en) * | 2016-07-28 | 2017-05-24 | 深圳前海达闼云端智能科技有限公司 | Access distribution method, device and system |
| CN106933675A (en) * | 2015-12-31 | 2017-07-07 | 华为技术有限公司 | A kind of method of process task, manager, server and system |
| CN107145406A (en) * | 2017-05-14 | 2017-09-08 | 四川盛世天成信息技术有限公司 | A kind of disaster-tolerant backup method and system based on Clustering |
| CN107172193A (en) * | 2017-06-20 | 2017-09-15 | 郑州云海信息技术有限公司 | A kind of load-balancing method and its device based on cluster |
| CN107257298A (en) * | 2017-07-27 | 2017-10-17 | 郑州云海信息技术有限公司 | A kind of fault handling method and device |
| CN107453909A (en) * | 2017-08-08 | 2017-12-08 | 北京盛华安信息技术有限公司 | The processing method of security incident |
| CN107592215A (en) * | 2017-08-31 | 2018-01-16 | 郑州云海信息技术有限公司 | A kind of method for managing message-oriented middleware |
| CN107615734A (en) * | 2015-05-26 | 2018-01-19 | 爱唯思有限公司 | For server failure transfer and the system and method for balancing the load |
| CN107682406A (en) * | 2017-09-08 | 2018-02-09 | 北京三快在线科技有限公司 | The method, apparatus and system of a kind of business processing |
| CN107689878A (en) * | 2016-08-04 | 2018-02-13 | 创盛视联数码科技（北京）有限公司 | TCP length connection SiteServer LBSs based on name scheduling |
| CN107689925A (en) * | 2017-09-28 | 2018-02-13 | 平安科技（深圳）有限公司 | Load balance optimization method and device based on cloud monitoring |
| CN107800551A (en) * | 2016-08-31 | 2018-03-13 | 北京优朋普乐科技有限公司 | Redis group systems and its method, the client for improving reliability |
| CN107948260A (en) * | 2017-11-15 | 2018-04-20 | 郑州云海信息技术有限公司 | Main monitoring node selecting method and device in a kind of distributed type assemblies |
| CN108763312A (en) * | 2018-04-26 | 2018-11-06 | 大连理工大学 | A kind of slave back end screening technique based on load |
| CN109474465A (en) * | 2018-11-13 | 2019-03-15 | 上海英方软件股份有限公司 | A kind of method and system of the high availability that can dynamically circulate based on server cluster |
| CN109634821A (en) * | 2018-11-13 | 2019-04-16 | 中科院微电子研究所昆山分所 | A kind of node availability detection method and device |
| CN109672728A (en) * | 2018-12-06 | 2019-04-23 | 航天恒星科技有限公司 | A kind of large-scale distributed clustered node information fast acquiring method |
| CN109728941A (en) * | 2018-12-20 | 2019-05-07 | 北京比新科技有限公司 | A kind of block chain leader election method and device thereof |
| CN109951331A (en) * | 2019-03-15 | 2019-06-28 | 北京百度网讯科技有限公司 | For sending the method, apparatus and computing cluster of information |
| CN110086726A (en) * | 2019-04-22 | 2019-08-02 | 航天云网科技发展有限责任公司 | A method of automatically switching Kubernetes host node |
| CN110708179A (en) * | 2018-07-10 | 2020-01-17 | 福建省天奕网络科技有限公司 | Block chain-based data communication bridging method and storage medium |
| CN111343254A (en) * | 2020-02-17 | 2020-06-26 | 天津卓朗科技发展有限公司 | Client connection method and device and electronic equipment |
| CN111400026A (en) * | 2019-11-15 | 2020-07-10 | 河海大学 | Distributed load balancing method based on master-slave backup technology |
| CN111541753A (en) * | 2020-04-16 | 2020-08-14 | 深圳市网心科技有限公司 | Distributed storage system, method, computer device and medium for block chain data |
| CN111565233A (en) * | 2020-05-28 | 2020-08-21 | 吉林亿联银行股份有限公司 | Data transmission method and device |
| CN111641522A (en) * | 2020-05-25 | 2020-09-08 | 杭州安恒信息技术股份有限公司 | Method, system and computer equipment for node switching |
| CN112190924A (en) * | 2020-12-04 | 2021-01-08 | 腾讯科技（深圳）有限公司 | Data disaster tolerance method, device and computer readable medium |
| CN113156803A (en) * | 2021-02-03 | 2021-07-23 | 南京华鹞信息科技有限公司 | Task-oriented unmanned aerial vehicle cluster resource management and fault-tolerant control method |
| CN113472566A (en) * | 2021-06-11 | 2021-10-01 | 北京市大数据中心 | Status monitoring method of union block chain and master node status monitoring system |
| CN113950165A (en) * | 2021-12-03 | 2022-01-18 | 中国电信集团系统集成有限责任公司 | Intelligent networking platform networking equipment connection method and device |
| CN114338727A (en) * | 2022-01-05 | 2022-04-12 | 中国建设银行股份有限公司 | Service cluster construction method and related device |
| CN114338696A (en) * | 2022-03-14 | 2022-04-12 | 北京奥星贝斯科技有限公司 | Method and device for distributed system |
| CN114978871A (en) * | 2022-04-08 | 2022-08-30 | 北京航天飞行控制中心 | Node switching method and node switching device of service system and electronic equipment |
| CN115001854A (en) * | 2022-07-18 | 2022-09-02 | 江苏艾盾网络科技有限公司 | Big data-based tracing-prevention server cluster management and control system and method |
| CN116112493A (en) * | 2023-02-09 | 2023-05-12 | 网易（杭州）网络有限公司 | Communication method, device, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1549978A (en) * | 2001-07-16 | 2004-11-24 | BEAϵͳ��˾ | Method and apparatus for session replication and failover |
| CN1925444A (en) * | 2006-09-14 | 2007-03-07 | 华为技术有限公司 | Method for establishing point-to-point collection in P2P network and nodes in P2P network |
| CN101521679A (en) * | 2009-04-03 | 2009-09-02 | 南京邮电大学 | Self-organizing method based on composite structured peer-to-peer network |
| CN101605092A (en) * | 2009-07-10 | 2009-12-16 | 浪潮电子信息产业股份有限公司 | A kind of content-based SiteServer LBS |
| CN102387218A (en) * | 2011-11-24 | 2012-03-21 | 浪潮电子信息产业股份有限公司 | Multimachine hot standby load balance system for computer |
| CN103117876A (en) * | 2013-01-24 | 2013-05-22 | 中兴通讯股份有限公司 | User state information synchronizing method and device |
-
2014
- 2014-09-26 CN CN201410512754.5A patent/CN104283948B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1549978A (en) * | 2001-07-16 | 2004-11-24 | BEAϵͳ��˾ | Method and apparatus for session replication and failover |
| CN1925444A (en) * | 2006-09-14 | 2007-03-07 | 华为技术有限公司 | Method for establishing point-to-point collection in P2P network and nodes in P2P network |
| US20090177772A1 (en) * | 2006-09-14 | 2009-07-09 | Huawei Technologies Co., Ltd. | Method, system and device for establishing a peer to peer connection in a p2p network |
| CN101521679A (en) * | 2009-04-03 | 2009-09-02 | 南京邮电大学 | Self-organizing method based on composite structured peer-to-peer network |
| CN101605092A (en) * | 2009-07-10 | 2009-12-16 | 浪潮电子信息产业股份有限公司 | A kind of content-based SiteServer LBS |
| CN102387218A (en) * | 2011-11-24 | 2012-03-21 | 浪潮电子信息产业股份有限公司 | Multimachine hot standby load balance system for computer |
| CN103117876A (en) * | 2013-01-24 | 2013-05-22 | 中兴通讯股份有限公司 | User state information synchronizing method and device |
Cited By (69)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105989279A (en) * | 2015-02-13 | 2016-10-05 | 上海通用识别技术研究所 | Domino dedicated password cracking system based on multi-hierarchy GPU (Graphic Process Unit) cluster |
| CN104768196A (en) * | 2015-04-21 | 2015-07-08 | 四川西结微波科技发展有限责任公司 | Cluster group seamless switching method |
| CN107615734A (en) * | 2015-05-26 | 2018-01-19 | 爱唯思有限公司 | For server failure transfer and the system and method for balancing the load |
| CN107615734B (en) * | 2015-05-26 | 2021-01-26 | 爱唯思有限公司 | System and method for server failover and load balancing |
| CN105007233A (en) * | 2015-07-13 | 2015-10-28 | 互联网域名系统北京市工程研究中心有限公司 | Method for distributing address based on DHCP (dynamic host configuration protocol) server cluster load |
| CN105007233B (en) * | 2015-07-13 | 2018-02-27 | 互联网域名系统北京市工程研究中心有限公司 | A kind of method that distribution address is loaded based on Dynamic Host Configuration Protocol server cluster |
| CN106453120A (en) * | 2015-08-05 | 2017-02-22 | 北京网御星云信息技术有限公司 | Dynamic clustering method and dynamic clustering system |
| CN106453120B (en) * | 2015-08-05 | 2019-06-07 | 北京网御星云信息技术有限公司 | A kind of dynamic cluster method and system |
| CN105338078B (en) * | 2015-10-26 | 2019-01-01 | 北京百度网讯科技有限公司 | Date storage method and device for storage system |
| CN105338078A (en) * | 2015-10-26 | 2016-02-17 | 北京百度网讯科技有限公司 | Data storage method and device used for storing system |
| CN105516245A (en) * | 2015-11-25 | 2016-04-20 | 国家计算机网络与信息安全管理中心 | Flow-based load balancing system and realization method thereof |
| 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 |
| CN105577759A (en) * | 2015-12-15 | 2016-05-11 | 东软熙康健康科技有限公司 | Server node allocation method and device |
| CN106933675A (en) * | 2015-12-31 | 2017-07-07 | 华为技术有限公司 | A kind of method of process task, manager, server and system |
| CN105430105B (en) * | 2016-01-05 | 2019-01-18 | 北京卓识达软件有限公司 | A kind of data processing method and system of car-mounted terminal |
| CN105430105A (en) * | 2016-01-05 | 2016-03-23 | 北京卓识达软件有限公司 | Data processing method and system for vehicle-mounted terminal |
| CN106027649A (en) * | 2016-05-20 | 2016-10-12 | 深圳市永兴元科技有限公司 | Data collection method of distributed data system and the distributed data system |
| CN106060123B (en) * | 2016-05-20 | 2019-08-16 | 深圳市永兴元科技股份有限公司 | The collecting method and distributed data system of distributed data system |
| CN106027649B (en) * | 2016-05-20 | 2019-08-16 | 深圳市永兴元科技股份有限公司 | The collecting method and distributed data system of distributed data system |
| CN106060123A (en) * | 2016-05-20 | 2016-10-26 | 深圳市永兴元科技有限公司 | Distributed data system data acquisition method and distributed data system |
| CN106716974B (en) * | 2016-07-28 | 2020-08-04 | 深圳前海达闼云端智能科技有限公司 | Access distribution method, device and system |
| CN106716974A (en) * | 2016-07-28 | 2017-05-24 | 深圳前海达闼云端智能科技有限公司 | Access distribution method, device and system |
| CN107689878A (en) * | 2016-08-04 | 2018-02-13 | 创盛视联数码科技（北京）有限公司 | TCP length connection SiteServer LBSs based on name scheduling |
| CN106331755B (en) * | 2016-08-17 | 2019-10-08 | 广州酷狗计算机科技有限公司 | The method and apparatus for handling Media Stream Caton |
| CN106331755A (en) * | 2016-08-17 | 2017-01-11 | 广州酷狗计算机科技有限公司 | Method and device for processing media stream blockage |
| CN107800551A (en) * | 2016-08-31 | 2018-03-13 | 北京优朋普乐科技有限公司 | Redis group systems and its method, the client for improving reliability |
| CN106549796A (en) * | 2016-09-27 | 2017-03-29 | 努比亚技术有限公司 | Resource control method and host node that a kind of firmware space is downloaded |
| CN107145406A (en) * | 2017-05-14 | 2017-09-08 | 四川盛世天成信息技术有限公司 | A kind of disaster-tolerant backup method and system based on Clustering |
| CN107172193A (en) * | 2017-06-20 | 2017-09-15 | 郑州云海信息技术有限公司 | A kind of load-balancing method and its device based on cluster |
| CN107257298A (en) * | 2017-07-27 | 2017-10-17 | 郑州云海信息技术有限公司 | A kind of fault handling method and device |
| CN107453909A (en) * | 2017-08-08 | 2017-12-08 | 北京盛华安信息技术有限公司 | The processing method of security incident |
| CN107592215A (en) * | 2017-08-31 | 2018-01-16 | 郑州云海信息技术有限公司 | A kind of method for managing message-oriented middleware |
| CN107682406B (en) * | 2017-09-08 | 2020-08-25 | 北京三快在线科技有限公司 | Method, device and system for processing service |
| CN107682406A (en) * | 2017-09-08 | 2018-02-09 | 北京三快在线科技有限公司 | The method, apparatus and system of a kind of business processing |
| US10992581B2 (en) | 2017-09-28 | 2021-04-27 | Ping An Technology (Shenzhen) Co., Ltd. | Methods and devices for optimizing load balancing based on cloud monitoring |
| CN107689925A (en) * | 2017-09-28 | 2018-02-13 | 平安科技（深圳）有限公司 | Load balance optimization method and device based on cloud monitoring |
| CN107689925B (en) * | 2017-09-28 | 2020-01-14 | 平安科技（深圳）有限公司 | Load balancing optimization method and device based on cloud monitoring |
| CN107948260A (en) * | 2017-11-15 | 2018-04-20 | 郑州云海信息技术有限公司 | Main monitoring node selecting method and device in a kind of distributed type assemblies |
| CN108763312A (en) * | 2018-04-26 | 2018-11-06 | 大连理工大学 | A kind of slave back end screening technique based on load |
| CN108763312B (en) * | 2018-04-26 | 2021-07-06 | 大连理工大学 | Slave data node screening method based on load |
| CN110708179A (en) * | 2018-07-10 | 2020-01-17 | 福建省天奕网络科技有限公司 | Block chain-based data communication bridging method and storage medium |
| CN109474465A (en) * | 2018-11-13 | 2019-03-15 | 上海英方软件股份有限公司 | A kind of method and system of the high availability that can dynamically circulate based on server cluster |
| CN109634821A (en) * | 2018-11-13 | 2019-04-16 | 中科院微电子研究所昆山分所 | A kind of node availability detection method and device |
| CN109672728A (en) * | 2018-12-06 | 2019-04-23 | 航天恒星科技有限公司 | A kind of large-scale distributed clustered node information fast acquiring method |
| CN109728941A (en) * | 2018-12-20 | 2019-05-07 | 北京比新科技有限公司 | A kind of block chain leader election method and device thereof |
| CN109951331B (en) * | 2019-03-15 | 2021-08-20 | 北京百度网讯科技有限公司 | Method, device and computing cluster for sending information |
| CN109951331A (en) * | 2019-03-15 | 2019-06-28 | 北京百度网讯科技有限公司 | For sending the method, apparatus and computing cluster of information |
| CN110086726A (en) * | 2019-04-22 | 2019-08-02 | 航天云网科技发展有限责任公司 | A method of automatically switching Kubernetes host node |
| CN111400026A (en) * | 2019-11-15 | 2020-07-10 | 河海大学 | Distributed load balancing method based on master-slave backup technology |
| CN111400026B (en) * | 2019-11-15 | 2023-02-28 | 河海大学 | Distributed load balancing method based on master-slave backup technology |
| CN111343254A (en) * | 2020-02-17 | 2020-06-26 | 天津卓朗科技发展有限公司 | Client connection method and device and electronic equipment |
| CN111343254B (en) * | 2020-02-17 | 2022-11-11 | 天津卓朗科技发展有限公司 | Client connection method and device and electronic equipment |
| CN111541753A (en) * | 2020-04-16 | 2020-08-14 | 深圳市网心科技有限公司 | Distributed storage system, method, computer device and medium for block chain data |
| CN111541753B (en) * | 2020-04-16 | 2024-02-27 | 深圳市迅雷网络技术有限公司 | Distributed storage system, method, computer device and medium for block chain data |
| CN111641522A (en) * | 2020-05-25 | 2020-09-08 | 杭州安恒信息技术股份有限公司 | Method, system and computer equipment for node switching |
| CN111565233A (en) * | 2020-05-28 | 2020-08-21 | 吉林亿联银行股份有限公司 | Data transmission method and device |
| CN112190924A (en) * | 2020-12-04 | 2021-01-08 | 腾讯科技（深圳）有限公司 | Data disaster tolerance method, device and computer readable medium |
| CN113156803A (en) * | 2021-02-03 | 2021-07-23 | 南京华鹞信息科技有限公司 | Task-oriented unmanned aerial vehicle cluster resource management and fault-tolerant control method |
| CN113472566A (en) * | 2021-06-11 | 2021-10-01 | 北京市大数据中心 | Status monitoring method of union block chain and master node status monitoring system |
| CN113950165B (en) * | 2021-12-03 | 2022-11-08 | 中电信数智科技有限公司 | Intelligent networking platform networking equipment connection method and device |
| CN113950165A (en) * | 2021-12-03 | 2022-01-18 | 中国电信集团系统集成有限责任公司 | Intelligent networking platform networking equipment connection method and device |
| CN114338727B (en) * | 2022-01-05 | 2024-03-12 | 中国建设银行股份有限公司 | Service cluster construction method and related device |
| CN114338727A (en) * | 2022-01-05 | 2022-04-12 | 中国建设银行股份有限公司 | Service cluster construction method and related device |
| WO2023173917A1 (en) * | 2022-03-14 | 2023-09-21 | 北京奥星贝斯科技有限公司 | Method and apparatus for distributed system |
| CN114338696A (en) * | 2022-03-14 | 2022-04-12 | 北京奥星贝斯科技有限公司 | Method and device for distributed system |
| CN114978871A (en) * | 2022-04-08 | 2022-08-30 | 北京航天飞行控制中心 | Node switching method and node switching device of service system and electronic equipment |
| CN114978871B (en) * | 2022-04-08 | 2023-10-24 | 北京航天飞行控制中心 | Node switching method and node switching device of service system and electronic equipment |
| CN115001854A (en) * | 2022-07-18 | 2022-09-02 | 江苏艾盾网络科技有限公司 | Big data-based tracing-prevention server cluster management and control system and method |
| CN116112493A (en) * | 2023-02-09 | 2023-05-12 | 网易（杭州）网络有限公司 | Communication method, device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104283948B (en) | 2018-12-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104283948A (en) | Server cluster system and load balancing implementation method thereof | |
| US9288113B2 (en) | Method and device for establishing structure of a communication network system | |
| CN102880475B (en) | Based on the real-time event disposal system of cloud computing and method in computer software | |
| CN107465721B (en) | Global load balancing method and system based on double-active architecture and scheduling server | |
| CN101288275B (en) | The performance characteristics of situation Network Based provides the method and apparatus of end-to-end high-quality service | |
| US10708175B2 (en) | Network address translation in networks using multiple NAT devices | |
| EP3629521B1 (en) | Rating method for service node, domain name system (dns) scheduling method and server | |
| CN102474903B (en) | Low latency mesh network | |
| CN104885431A (en) | Content based traffic engineering in software defined information centric networks | |
| CN106340176A (en) | Intelligent electricity meter information sharing method, intelligent electricity meter and acquisition router | |
| CN104270291A (en) | Content delivery network (CDN) quality monitoring method | |
| CN103207841A (en) | Method and device for data reading and writing on basis of key-value buffer | |
| CN104243196A (en) | Virtual network mapping protection method and system under SDN architecture | |
| CN103259744A (en) | Method for mapping mobile virtual network based on clustering | |
| CN106658559B (en) | A kind of Mobile QoS keeping method based on context-prediction | |
| JP5869041B2 (en) | Method for mapping network topology request to physical network, computer program product, mobile communication system, and network configuration platform | |
| CN106713378B (en) | Method and system for providing service by multiple application servers | |
| CN102984184A (en) | A method and a device for service load balancing for a distributed system | |
| CN112311896B (en) | Health examination method, device, equipment and computer readable storage medium | |
| CN107872368A (en) | Detection method, device and the terminal of gateway accessibility in a kind of network node cluster | |
| CN106533973A (en) | Method and system for distributing service message, and equipment | |
| CN105391651A (en) | Virtual optical network multilayer resource convergence method and system | |
| CN102882979B (en) | Data acquisition based on cloud computing system and the system and method collecting shunting | |
| CN106357723A (en) | Synchronous system and method for multi-cluster information caching based on cloud host | |
| CN104579729B (en) | The notification method and device of CGN single board default |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |