CN103051551B - A kind of distributed system and its automatic maintenance method - Google Patents
A kind of distributed system and its automatic maintenance method Download PDFInfo
- Publication number
- CN103051551B CN103051551B CN201110310214.5A CN201110310214A CN103051551B CN 103051551 B CN103051551 B CN 103051551B CN 201110310214 A CN201110310214 A CN 201110310214A CN 103051551 B CN103051551 B CN 103051551B
- Authority
- CN
- China
- Prior art keywords
- service
- node
- message
- log
- request
- 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.)
- Active
Links
Landscapes
- Multi Processors (AREA)
Abstract
The invention discloses a kind of distributed system and its automatic maintenance method, the above method comprises the following steps:After node starts service, regularly service registration request is sent to service register center;Above-mentioned service register center is safeguarded, and provide information service to above-mentioned node or its associated nodes according to the service registration request received to local information.The distributed system scale of the present invention can be with Dynamic contraction, and node can automatically initiate service registry after starting, without manual configuration;Network relation is automatically formed by call relation between node and is dynamically adapted, network structure safeguards simple that construction cost is relatively low without advance planning, in addition present system.
Description
Technical field
The present invention relates to the distributed computing technology in field of cloud calculation, more particularly to a kind of distributed system and its automatic maintenance
Method.
Background technology
It is growing with internet, the user of conventional telecommunications net and application, and both fusions (turn into mobile interchange
Net) trend continuous reinforcement, also constantly increased by the data traffic and access request of server.This deployment to server
With safeguarding, the scalability of operation software performance thereon and whole system propose higher and higher demand.For this
A little demands, prior art propose the distributed computing system of some replies.When server is excessively busy, by increasing system
Load-balancing device and number of servers, or lift its processing by lifting the hardware and software performance of individual server
Ability.Because server cost is too high, performance growth potential is limited, its performance is lifted be not main flow solution party
Case.
Fig. 1 is a kind of distributed system that prior art proposes, as shown in figure 1, this system passes through load-balancing device
One group of server is supported, load-balancing device is placed on the front end of Servers-all, outwards the interface of unified exposure service.Load
Balancing equipment is probably hardware, such as the fourth layer machine;It is also likely to be software, such as some reverse proxy softwares.This system is general
Applied on the website of internet, its design philosophy is that numerous WEB servers are formed into cluster by load-balancing device, so as to
Provide a user single HTML services.This system when servicing (application) and increasing, it is necessary to set up new load-balancing device,
Because each service is required for one group of server to support, and each group of server all uses a load-balancing device, this gesture
Steeply rising for lower deployment cost must be caused, load-balancing device can also can't bear the heavy load because of the substantial increase of access request.Separately
Outside, when internal system increases one group of service newly, client computer can not automatic sensing these newly-increased services server address, it is necessary to
Access request could be smoothly sent to new server after a series of manual configurations.There are hundreds and thousands of services in the modern times
It is extremely difficult, poorly efficient that such manual configuration is done by the data center of device.
Therefore, there is the shortcomings of lower deployment cost is too high, later maintenance is difficult, system robustness is poor in this distributed system,
The system can not solve the problems such as server performance bottleneck caused by service and user's growth.
Fig. 2 is another distributed system that prior art proposes, as shown in Fig. 2 this system passes through ESB
(Enterprise Service Bus, ESB) server supports one group of service, and all access requests are required for leading to
Cross the forwarding of ESB servers.ESB servers are based on SOA (Service-Oriented Architecture, Service-Oriented Architecture Based)
Thought, there is provided load-balancing mechanism, newly-increased service are also transparent to client computer.However, ESB servers assume responsibility in itself
Too many function, single point failure hidden danger and unit performance bottleneck be present.On the other hand, a kind of service or clothes are increased newly on such a system
Be engaged in node when, it is necessary on ESB servers manual registration, the problem of difficult in maintenance, scalability is poor be present.ESB servers
Another limitation be that generally it directs out exposure service, can not mutually call between internal services, therefore it is general
Needed applied to enterprise by internal numerous system combinations and the situation of standards service is externally provided, it is extensive to be not particularly suited for structure
Distributed system.
The content of the invention
It is an object of the present invention to provide a kind of distributed system and its automatic maintenance method, to solve existing distribution
Lower deployment cost present in system is too high, performance bottleneck is serious, later maintenance is difficult, system robustness and retractility difference etc. are asked
Topic.
The present invention solves above-mentioned technical problem by the following technical programs:
A kind of automatic maintenance method of large scale distributed system, the above method comprise the following steps:
After node starts service, regularly service registration request is sent to service register center;
Above-mentioned service register center safeguarded to local information according to the service registration request received, and to above-mentioned section
Point or its associated nodes provide information service.
According to the above method, above-mentioned node includes providing the node of service and the node of transfer service, above-mentioned service registry
Include the log-on message of above-mentioned node in request, above-mentioned log-on message includes the IP address of above-mentioned node, port numbers and above-mentioned
Node can transfer or offer service name.
Above-mentioned service register center is safeguarded using following methods to local information:
Log-on message in being asked according to above-mentioned registration service, judges whether above-mentioned node is to register first, if so, then will
Above-mentioned log-on message is saved in local node chained list, otherwise by above-mentioned node in above-mentioned log-on message and above-mentioned node linked list
Old log-on message is compared, and updates above-mentioned node linked list with above-mentioned log-on message if both are inconsistent.
Above-mentioned service register center above-mentioned log-on message is saved in local node chained list or with above-mentioned log-on message more
It is further comprising the steps of after new above-mentioned node linked list:
Judge whether above-mentioned node is the node for providing service;
If so, the log-on message of above-mentioned node is then sent to the associated nodes using above-mentioned node serve, otherwise, upwards
State node and its log-on message for wanting the destination node belonging to the service of transfer is provided.
The above method also includes the step of following above-mentioned node calls service:
Above-mentioned node sends service invocation request to above-mentioned service register center first;
After above-mentioned service register center receives above-mentioned call request, transfer or the institute of above-mentioned node required service will can be provided
The log-on message for having destination node returns to above-mentioned node;
Above-mentioned node preserves the log-on message of destination node, and obtains the loading condition of each destination node, then to load
Minimum destination node sends service invocation request;
After the minimum destination node of above-mentioned load receives above-mentioned call request, the service that above-mentioned node needs is returned to
State node.
According to the above method, after above-mentioned service register center receives above-mentioned call request, transfer or above-mentioned section can be provided
The log-on message of all destination nodes of point required service, which returns to above-mentioned node step, to be included:
The service that above-mentioned node to be called is saved in local above-mentioned node linked list by above-mentioned service register center;
The above-mentioned service of node transfer is determined whether according to local above-mentioned node linked list, if so, then can the above-mentioned clothes of transfer
The log-on message of all destination nodes of business is sent to above-mentioned node, otherwise, it is determined whether there is node to provide above-mentioned service;
If so, the log-on message for all destination nodes that above-mentioned service can be provided then is sent into above-mentioned node, otherwise return
Return the information of service request failure.
Above-mentioned node obtains the loading condition of each above-mentioned destination node using following methods:
According to the log-on message of above-mentioned destination node, the heartbeat message of itself is regularly sent to each destination node;
After each destination node receives heartbeat message, the heartbeat message of itself is returned into above-mentioned node;
Above-mentioned node obtains its loading condition according to the heartbeat message of each destination node.
Above-mentioned heartbeat message includes the IP address and load data of node.
After above-mentioned node saves the log-on message of destination node, if next time asks above-mentioned service again, skip to above-mentioned
Service register center sends the step of service invocation request, directly sends service invocation request to the minimum destination node of load,
Untill asking to fail.
The automatic maintenance method of above-mentioned distributed system also includes:When above-mentioned node does not send server registration request regularly
When, above-mentioned service register center deletes above-mentioned node in the log-on message locally preserved.
The present invention also uses following technical scheme:
A kind of distributed system, including node and service register center;
Above-mentioned node, for after service is started, regularly sending service registration request to service register center;
Above-mentioned service register center, the service registration request received for basis, is safeguarded to local information, and upwards
State node or its associated nodes provides information service.
Above-mentioned node includes information registering module, service request module, optional service providing module and service and receives mould
Block;
Above- mentioned information Registering modules, service registration request is sent to above-mentioned service register center for timing;
Above-mentioned service request module, for sending service invocation request to above-mentioned service register center or other nodes;
Above-mentioned service providing module, after receiving service invocation request, the service that itself is provided returns to accordingly
Node;
Above-mentioned service receiving module, for receiving the service of other nodes offer.
Above-mentioned service register center includes maintenance of information module, Service Notification module and service response module;
Above- mentioned information maintenance module, for receiving the registration request of above-mentioned node, and according to above-mentioned registration request to local
The nodal information of preservation is safeguarded;
Above-mentioned Service Notification module, for newly-increased or renewal nodal information to be sent into other associated nodes, or upwards
State newly-increased or renewal node it is provided and want nodal information belonging to the service of transfer;
Above-mentioned service response module, for after the service invocation request that above-mentioned node is sent is received, transfer or it can carry
Above-mentioned node is returned to for the nodal information of respective service.
Compared with prior art, the present invention has following advantageous effects:
1) Automatic service discovery:Different, this hair of manual setting is needed when new demand servicing is registered from traditional distributed system
Bright service node can register its service type after starting from trend service register center;Other nodes need to ask this service
When, it can obtain servicing position by service register center.This characteristic has the advantage that network structure is simple, communication
Link can be established automatically according to the call relation between node, the network structure (legacy system being not required between advance planning node
Advance planning is generally wanted, also can not dynamically be adjusted), this facilitate that dynamic increase in the data center or deletion of node;
2) automatic equalization loads:It is minimum to automatically select load in the service that request target node provides for node of the present invention
Node establish communication link, concurrently send service invocation request.So that in the case where not introducing conventional load balancer device,
The present invention can also realize the load balancing between node.This not only saves cost, while also eliminates single load equalizer
Single point failure hidden danger and single-point performance bottleneck, system robustness it is stronger;
3) scale is flexible convenient:It is loose couplings between node of the present invention, any node failure, even in service registry
Heart failure, continued to run with all without influence system, increase or deletion of node are only to promote service request amount to weigh among the nodes
New distribution;Therefore, the scale of present system is stretched very convenient, is advantageous to build large-scale distributed computing system;
4) service relatively independent:Node of the present invention is equal when sending registration request, request call service, obtaining loading condition
The service (software) that node itself is run it is not related to, i.e. software is not involved in these functions in itself, and these functions are by node bottom
The communications framework of layer is automatically performed, and software need to only focus on the realization of service logic.It reduce the clothes of distributed computing system
Business development difficulty.
Brief description of the drawings
Accompanying drawing described herein is used for providing a further understanding of the present invention, forms the part of the present invention, this hair
Bright schematic description and description is used to explain the present invention, does not form inappropriate limitation of the present invention.In the accompanying drawings:
Fig. 1 is a kind of structural representation for distributed system that prior art proposes;
Fig. 2 is the structural representation for another distributed system that prior art proposes;
Fig. 3 is the structural representation of distributed system preferred embodiment of the present invention;
The flow chart of the automatic maintenance method preferred embodiment of Fig. 4 distributed systems of the present invention;
Fig. 5 is the module frame chart of distributed system preferred embodiment of the present invention.
Embodiment
In order that technical problems, technical solutions and advantages to be solved are clearer, clear, tie below
Drawings and examples are closed, the present invention will be described in further detail.It should be appreciated that specific embodiment described herein is only used
To explain the present invention, it is not intended to limit the present invention.
Fig. 3 is the structural representation of distributed system preferred embodiment of the present invention.As shown in figure 3, in the embodiment of the present invention
Distributed system there are multiple nodes, including provide the node (service node) of service and node (the middle trochanter of transfer service
Point), the hardware of each node can be less expensive computer, but rely on the powerful retractility of system, on the whole system
Very powerful computing capability can be provided.Computer bottom can be passed through between node and node, node and service register center
Layer communications framework establishes communication link, such as TCP communication link, and information or the function of service are persistently sent to other side to realize.
One node can establish communication link after the log-on message of other nodes is obtained, and be sent to service and adjust
With request.Service register center forms as the core of system, saves the log-on messages of all nodes, including all nodes
IP address, port numbers and node provide or the service of transfer.Each service can be distributed in multiple service nodes, i.e., one group
Service node can provide same service type.Transit node can be with transfer service, but itself does not carry (offer) clothes
Business, therefore transit node is optional, but transit node can simplify the network topology structure of system.For example, section is serviced in Fig. 3
When point 1 will call service B, because service node 4 and 5 can provide this service, therefore it must first obtain and take to service register center
The log-on message of business node 4 and 5, then judges their loading condition, then asks to load and taken needed for small service node return
Business.And if be provided with transit node 1, service node 1 only needs service invocation request being sent to transit node 1.
It is worth noting that, the present embodiment service node 1 obtain service node 4 and 5 log-on message after, can timing to
Service node 4 and 5 sends heartbeat message, and thus service node 1 can not only obtain the loading condition of other side, can also not receive
To the information that other side failure/exit is grasped during other side's heartbeat message.For example, if service node 1 does not receive the heartbeat of service node 5
During information, service node 1 judges the failure of service node 5/exit, and now service invocation request is no longer sent to service node by it
5, untill service register center notifies that its service node 5 restarts.
It is further to note that the distributed system scale of the present embodiment can be with dynamic retractility, Fig. 3 is schematically
Give A, B two groups of services, actually the present invention can support infinite number of service, while the present invention can as needed at any time
Increase new service or transit node with lifting system performance, partial service node can also be closed when service request tails off.
The addition of node and exit all be it is dynamic, without by hand be configured because the present invention by service register center from
It is dynamic to safeguard each node and its log-on message, it can be initiated to register to service register center when new node adds, other associated nodes
The information of new node can be obtained rapidly by service register center, just share partial service call request on new node.
It is regularly to send log-on message to service register center additionally, due to node, therefore when some node exits, service note
Volume information can perceive this situation rapidly, and by the nodal information from local deletion, and other associated nodes are believed by heartbeat
What breath received stops breaking, and can also be immediately detected this situation, new service invocation request will not re-send to the node.
The structure to distributed system of the present invention and its maintenance (operation) mechanism are described in detail on the whole above,
The automatic maintenance method of distributed system provided by the invention is described in detail below in conjunction with accompanying drawing 4.
The flow chart of the automatic maintenance method preferred embodiment of Fig. 4 distributed systems of the present invention, as shown in figure 4, of the invention
The method of preferred embodiment comprises the following steps:
Step S1:After node starts service, regularly service registration request is sent to service register center;
After a node starts service, it starts timing and sends service registration request, above-mentioned note to service register center
Volume request include node IP address, port numbers and node can transfer or offer service name, its interior joint can transfer or
The service of offer can be one or more.By regularly sending above-mentioned registration request, service register center can understand in time
The running situation of node, and know that node breaks down or logged off when not receiving the registration request that node timing is sent
Information.
Step S2:The service registration request that service register center receiving node is sent;
Step S3:Service register center is safeguarded according to the service registration request received to local information;
Service register center is safeguarded to local information by the following method:
Log-on message in being asked according to above-mentioned registration service, judges whether above-mentioned node is to register first, if so, then will
Above-mentioned log-on message is saved in local node chained list, otherwise by above-mentioned node in above-mentioned log-on message and above-mentioned node linked list
Old log-on message is compared, and updates above-mentioned node linked list with above-mentioned log-on message if both are inconsistent.
Step S4:Service register center provides information service to above-mentioned node or its associated nodes;
Service register center provides information service using following methods to above-mentioned node or its associated nodes:
Judge whether above-mentioned node is the node for providing service:
If the log-on message of above-mentioned node is then sent to the associated nodes using above-mentioned node serve, otherwise to above-mentioned
Node provides its log-on message for wanting the destination node belonging to the service of transfer.
By step S4, when thering is service node to add or updating, it is necessary to be passed through using other nodes of the node serve
Service register center can know the node situation rapidly, and service request is shared on the node.Add when there is transit node
When entering or updating, transit node can also know that it wants the nodal information belonging to the service of transfer rapidly by service register center,
Thus when it receives the service invocation request of node, service node can be forwarded a request to.
Step S5:Above-mentioned node sends service invocation request to service register center;
When above-mentioned node needs to use the service that some node A provides, and above-mentioned node is local records node A's
During log-on message, service invocation request, acquisition request node A log-on message, including IP can be sent to service register center
Address, service port number etc., so as to establish communication linkage with node A, and obtain the service of its offer.
Step S6:Service register center receives above-mentioned request, transfer or will can provide the target of above-mentioned node required service
The log-on message of node returns to above-mentioned node;
After service register center receives above-mentioned request, the service that above-mentioned node needs to call is stored in local first, with
Just when subsequently having new node to add and its required service can be provided, the log-on message of new node is actively returned into above-mentioned section
Point;
Then the service that node can be needed for the above-mentioned node of transfer is determined whether according to the node linked list information of local record,
If so, then by can the log-on message of destination node of the above-mentioned node required service of transfer be sent to above-mentioned node, otherwise searching is
It is no to there is node to provide the service needed for above-mentioned node;
If so, the log-on message for all destination nodes that above-mentioned node required service can be provided then is sent to above-mentioned section
Point, if otherwise returning to the information of service request failure.
For example, the service node 1 in Fig. 3 needs the unknown service B in request call address of service, then service node 1 passes through
Bottom communication framework sends service B service invocation request to service register center, after service register center receives request, first
Service node 1 is needed to use into service B information record into local node chained list, whether then inquire about has node can transfer clothes
Be engaged in B, as a result finds that transit node 1 can service B with transfer, the log-on message of transit node 1 then is sent into service node 1.
Assuming that the failure of transit node 1 or log off, then service register center inquiry find no node can transfer service B, then look into
Whether inquiry has node to provide service B, as a result finds that presence service node 4 and service node 5 can provide directly service B, now takes
The log-on message of service node 4 and service node 5 is sent to service node 1 by business registration center.
Step S7:Above-mentioned node preserves the log-on message of destination node, obtains the loading condition of each destination node, and to negative
Carry minimum destination node and send service invocation request;
Above-mentioned node receive can transfer or after the log-on message of all destination nodes of itself required service is provided, by this
The log-on message of a little destination nodes is stored in local, so as to call these destination nodes to provide again next time service when, skip
The step of sending service invocation request to service register center, directly sending service call to the minimum destination node of load please
Ask, until just sending service invocation request to service register center again during certain request failure.
Above-mentioned node obtains the loading condition of each destination node using following methods:
According to the log-on message of each destination node, the heartbeat message of itself is regularly sent to each destination node;It is each
After destination node receives the heartbeat message of above-mentioned node, respective heartbeat message is returned into above-mentioned node;Above-mentioned node receives
The heartbeat message of each destination node, and therefrom obtain the loading condition of each destination node.
Because the heartbeat message includes the IP address and load data of node itself, thus above-mentioned node by with target
Node is mutually made up one's mind hop-information, and can obtain the loading condition of other side.Above-mentioned node sends the heart by timing to each destination node
Hop-information, the loading condition of other side can be not only obtained, whether can broken down or log off with master goal node
More information.For example, when above-mentioned node does not receive the heartbeat message of some destination node return, it can be determined that the target section
Point has broken down or logged off, and hereafter service registration request will no longer issue the destination node.
Such as assume in previous step S6, when the service node 1 in Fig. 3 is to service register center request call service B,
The log-on message of service node 4 and service node 5 is have received, then in the step s 7, service node 1 is first by the He of service node 4
The log-on message of service node 5 is stored in local, then heartbeat message is sent to service node 4 and service node 5 respectively, to obtain
Their loading condition is taken, finally selection establishes communication link with loading a less node, and sends service call to it
Request.
Step S8:After the minimum node of load receives service invocation request, the service needed for above-mentioned node is returned to
State node;
When the minimum node of above-mentioned load is transit node, after it receives above-mentioned request, above-mentioned request is transmitted to can
The node of above-mentioned service is directly provided, and request results are returned into above-mentioned node;When the minimum node of above-mentioned load is service
During node, after it receives above-mentioned request, the service needed for above-mentioned node is directly returned into above-mentioned node.
Step S9:When above-mentioned node does not send server registration request regularly, service register center deletes above-mentioned node
In the log-on message locally preserved.
Fig. 5 is the module frame chart of distributed system preferred embodiment of the present invention, as shown in figure 5, the preferred embodiment of the present invention
Distributed system include:Node 51 and service register center 52;
Above-mentioned node 51, for after service is started, regularly sending service registration request to service register center 52;
Above-mentioned service register center 52, the service registration request received for basis, is safeguarded to local information, and to
Above-mentioned node 51 or its associated nodes provide information service.
Above-mentioned node 51 include information registering module 511, service request module 512, optional service providing module 513,
And service receiving module 514;
Above- mentioned information Registering modules 511, service registration request is sent to above-mentioned service register center 52 for timing;
Above-mentioned service request module 512, please for sending service call to above-mentioned service register center 52 or other nodes
Ask;
Above-mentioned service providing module 513, for after service invocation request is received, the service that itself is provided to return to phase
The node answered;
Above-mentioned service receiving module 514, for receiving the service of other nodes offer.
Above-mentioned service register center 52 includes maintenance of information module 521, Service Notification module 522 and service response module
523;
Above- mentioned information maintenance module 521, for receiving the registration request of above-mentioned node 51, and according to above-mentioned registration request pair
The nodal information locally preserved is safeguarded;
Above-mentioned Service Notification module 522, for newly-increased or renewal nodal information to be sent into other associated nodes, or to
Above-mentioned newly-increased or renewal node provides it and wants nodal information belonging to the service of transfer;
Above-mentioned service response module 523, can transfer for after the service invocation request that above-mentioned node 51 is sent is received
Or the destination node information of offer respective service returns to above-mentioned node 51.
The preferred embodiments of the present invention have shown and described in described above, but as previously described, it should be understood that the present invention is not
Form disclosed herein is confined to, is not to be taken as the exclusion to other embodiment, and available for various other combinations, modification
And environment, and can be carried out in the scope of the invention is set forth herein by the technology or knowledge of above-mentioned teaching or association area
Change., then all should be in institute of the present invention and the change and change that those skilled in the art are carried out do not depart from the spirit and scope of the present invention
In attached scope of the claims.
Claims (12)
1. a kind of automatic maintenance method of distributed system, it is characterised in that the described method comprises the following steps:
After node starts service, regularly service registration request is sent to service register center;
The service register center safeguarded to local information according to the service registration request received, and to the node or
Its associated nodes provides information service;
The node includes providing the node of service and the node of transfer service;
The service node can register its service type after starting from trend service register center;
Methods described also includes the step of node calls service:
The node sends service invocation request to the service register center first;
After the service register center receives the call request, transfer or all mesh of the node required service will can be provided
The log-on message of mark node returns to the node;
The node preserves the log-on message of destination node, and obtains the loading condition of each destination node, then minimum to load
Destination node send service invocation request;
After the minimum destination node of the load receives the call request, the service that the node needs is returned into the section
Point.
2. according to the method for claim 1, it is characterised in that:The registration of the node is included in the service registration request
Information, IP address, port numbers and the node of the log-on message including the node can transfer or offer Service name
Claim.
3. according to the method for claim 2, it is characterised in that the service register center is using following methods to local letter
Breath is safeguarded:
Log-on message in being asked according to the registration service, judges whether the node is to register first, if so, then by described in
Log-on message is saved in local node chained list, otherwise by the old note of node described in the log-on message and the node linked list
Volume information is compared, and updates the node linked list with the log-on message if both are inconsistent.
4. according to the method for claim 3, it is characterised in that the log-on message is saved in by the service register center
Updated in local node chained list or with the log-on message further comprising the steps of after the node linked list:
Judge whether the node is the node for providing service;
If so, the log-on message of the node is then sent to the associated nodes using the node serve, otherwise, to the section
Point provides its log-on message for wanting the destination node belonging to the service of transfer.
5. according to the method for claim 1, it is characterised in that after the service register center receives the call request,
Transfer or the log-on message of all destination nodes of the node required service will can be provided return to the node step and include:
The service that the node to be called is saved in the local node linked list by the service register center;
The service according to the local node linked list determines whether node transfer, if so, be able to will then service described in transfer
The log-on message of all destination nodes is sent to the node, otherwise, it is determined whether there is node to provide the service;
If so, the log-on message for all destination nodes that the service can be provided then is sent into the node, clothes are otherwise returned
The information of business request failure.
6. according to the method for claim 1, it is characterised in that the node obtains each target using following methods
The loading condition of node:
According to the log-on message of the destination node, the heartbeat message of itself is regularly sent to each destination node;
After the destination node receives heartbeat message, the heartbeat message of itself is returned into the node;
The node obtains its loading condition according to the heartbeat message of destination node.
7. according to the method for claim 6, it is characterised in that:The heartbeat message includes the IP address and load number of node
According to.
8. according to the method for claim 1, it is characterised in that:After the node saves the log-on message of destination node,
If when asking the service again, skip next time to the service register center send service invocation request the step of, directly to negative
Carry minimum destination node and send service invocation request, untill asking to fail.
9. according to the method for claim 1, it is characterised in that methods described also includes:When the node is not sent regularly
During service registration request, the service register center deletes the node in the log-on message locally preserved.
10. a kind of distributed system, including node, it is characterised in that also include:Service register center;
The node, for after service is started, regularly sending service registration request to service register center;
The service register center, for according to the service registration request that receives, being safeguarded to local information, and to the section
Point or its associated nodes provide information service;
The node includes providing the node of service and the node of transfer service;
The service node can register its service type after starting from trend service register center;
The node, it is additionally operable to call service, the node sends service invocation request to the service register center first;
The service register center, it is additionally operable to after receiving the call request, transfer or the node required service can be provided
The log-on messages of all destination nodes return to the node;
The node, is additionally operable to preserve the log-on message of destination node, and obtains the loading condition of each destination node, then to negative
Carry minimum destination node and send service invocation request;
After the minimum destination node of the load receives the call request, the service that the node needs is returned into the section
Point.
11. system according to claim 10, it is characterised in that:The node includes information registering module, service request
Module, optional service providing module and service receiving module;
Described information Registering modules, service registration request is sent to the service register center for timing;
The service request module, for sending service invocation request to the service register center or other nodes;
The service providing module, the service for after service invocation request is received, itself to be provided return to corresponding section
Point;
The service receiving module, for receiving the service of other nodes offer.
12. system according to claim 10, it is characterised in that:The service register center include maintenance of information module,
Service Notification module and service response module;
Described information maintenance module, for receiving the registration request of the node, and according to the registration request to local preservation
Nodal information safeguarded;
The Service Notification module, for newly-increased or renewal nodal information to be sent into other associated nodes, or to described new
Increase or the node of renewal provides it and wants nodal information belonging to the service of transfer;
The service response module, for after the service invocation request that the node is sent is received, transfer or phase can be provided
The nodal information that should be serviced returns to the node.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110310214.5A CN103051551B (en) | 2011-10-13 | 2011-10-13 | A kind of distributed system and its automatic maintenance method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110310214.5A CN103051551B (en) | 2011-10-13 | 2011-10-13 | A kind of distributed system and its automatic maintenance method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103051551A CN103051551A (en) | 2013-04-17 |
CN103051551B true CN103051551B (en) | 2017-12-19 |
Family
ID=48064056
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201110310214.5A Active CN103051551B (en) | 2011-10-13 | 2011-10-13 | A kind of distributed system and its automatic maintenance method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103051551B (en) |
Families Citing this family (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104426936A (en) * | 2013-08-22 | 2015-03-18 | 中兴通讯股份有限公司 | Load balancing method and system |
CN103457870B (en) * | 2013-09-25 | 2016-06-29 | 武汉理工大学 | For the load balancing of distributed system and high availability subsystem and method |
CN103577256B (en) * | 2013-11-21 | 2017-03-15 | 五八同城信息技术有限公司 | Distributed timing task scheduling system |
CN103780434A (en) * | 2014-02-18 | 2014-05-07 | 成都致云科技有限公司 | Unattended operation automatic networking method of large-type distribution type system |
CN103986748B (en) * | 2014-04-22 | 2019-01-25 | 数贸科技(北京)有限公司 | The method and apparatus of the serviceization of realization |
CN105407161B (en) * | 2015-11-30 | 2019-06-14 | 北京奇艺世纪科技有限公司 | Application service management method, system and application service find method, system |
CN106874155A (en) * | 2015-12-14 | 2017-06-20 | 中国航空工业第六八研究所 | A kind of dynamic monitoring and controlling method of embedded software developing stage for target machine |
CN105915583B (en) * | 2016-03-28 | 2020-05-26 | 联想(北京)有限公司 | Method for starting service cluster and service cluster |
CN106210120B (en) * | 2016-07-29 | 2019-07-09 | 广州南方仕通网络科技有限公司 | A kind of recommended method and its device of server |
CN106453539B (en) * | 2016-09-28 | 2019-04-05 | 广州市百果园网络科技有限公司 | A kind of method and system of processing Node registry notice |
CN106790084A (en) * | 2016-12-22 | 2017-05-31 | 中国科学院电子学研究所苏州研究院 | A kind of heterogeneous resource integrated framework and its integrated approach based on ICE middlewares |
CN109218355B (en) * | 2017-06-30 | 2021-06-15 | 华为技术有限公司 | Load balancing engine, client, distributed computing system and load balancing method |
CN107147466B (en) * | 2017-07-20 | 2019-03-12 | 郑州云海信息技术有限公司 | A kind of control method and system of system time consistency |
CN109040311B (en) * | 2018-09-17 | 2021-07-20 | 中国联合网络通信集团有限公司 | Service information push processing method and device |
CN109120717B (en) * | 2018-09-26 | 2023-07-25 | 中国平安人寿保险股份有限公司 | Reverse proxy method, device, storage medium and distributed system |
CN109829093A (en) * | 2018-12-29 | 2019-05-31 | 深圳云天励飞技术有限公司 | A kind of method and device of search |
CN110971702A (en) * | 2019-12-10 | 2020-04-07 | 中国建设银行股份有限公司 | Service calling method and device, computer equipment and storage medium |
CN111262916B (en) * | 2020-01-10 | 2022-11-08 | 深圳市网心科技有限公司 | Data distributed storage method and device, storage node and storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
CN101378327A (en) * | 2007-08-29 | 2009-03-04 | 中国移动通信集团公司 | Communication network system and method for processing communication network business |
CN101997759A (en) * | 2009-08-10 | 2011-03-30 | 中兴通讯股份有限公司 | Service implementation method and service system |
-
2011
- 2011-10-13 CN CN201110310214.5A patent/CN103051551B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
CN101378327A (en) * | 2007-08-29 | 2009-03-04 | 中国移动通信集团公司 | Communication network system and method for processing communication network business |
CN101997759A (en) * | 2009-08-10 | 2011-03-30 | 中兴通讯股份有限公司 | Service implementation method and service system |
Also Published As
Publication number | Publication date |
---|---|
CN103051551A (en) | 2013-04-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103051551B (en) | A kind of distributed system and its automatic maintenance method | |
US20170310596A1 (en) | Load distribution in data networks | |
JP4616159B2 (en) | Cluster system, load balancer, node transfer method, and node transfer program | |
CN101809966B (en) | Method and apparatus for establishing and managing diameter associations | |
EP1744519B1 (en) | Information exchange system, management server, and method for reducing the network load | |
CN113596110B (en) | Cloud primary micro-service platform oriented to heterogeneous cloud | |
US7233999B2 (en) | System and method for delivering last mile computing over light from a plurality of network edge locations | |
US20140280606A1 (en) | Method and Apparatus for Content Management | |
JP2020537449A (en) | Service registration in communication network | |
CN109189856A (en) | Distributed data base service management, device, server and storage medium | |
CN108833462A (en) | A kind of system and method found from registration service towards micro services | |
US20090319686A1 (en) | Communication route selecting method and apparatus | |
WO2004036344A2 (en) | System and method for the optimization of database | |
AU2006203017A1 (en) | Information exchange system and management server, terminal unit, and method for reducing network load used in the same | |
CN108632299A (en) | Enhance method, apparatus, electronic equipment and the storage medium of registration center's availability | |
EP1180886B1 (en) | Load balancing method and system | |
US10659427B1 (en) | Call processing continuity within a cloud network | |
CN107888551A (en) | A kind of remote service calling method, device and system | |
CN112351077B (en) | Application service operation method, system, device and storage medium | |
US20150006622A1 (en) | Web contents transmission method and apparatus | |
US9806965B2 (en) | Automatic user redundancy determination | |
CN109120556B (en) | A kind of method and system of cloud host access object storage server | |
JP2002351760A (en) | Device and method for decentralizing server load, and program making computer implement the same method | |
CN100484014C (en) | Distributed cluster service management system and service management method in intelligent network | |
CN102546712B (en) | Message transmission method, equipment and system based on distributed service network |
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 |