CN103475566A - Real-time message exchange platform and distributed cluster establishment method - Google Patents
Real-time message exchange platform and distributed cluster establishment method Download PDFInfo
- Publication number
- CN103475566A CN103475566A CN2013102877013A CN201310287701A CN103475566A CN 103475566 A CN103475566 A CN 103475566A CN 2013102877013 A CN2013102877013 A CN 2013102877013A CN 201310287701 A CN201310287701 A CN 201310287701A CN 103475566 A CN103475566 A CN 103475566A
- Authority
- CN
- China
- Prior art keywords
- node
- message
- server
- gateway
- ordinary
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Computer And Data Communications (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a real-time message exchange platform and a distributed cluster establishment method. The platform includes an application server cluster and a data server cluster. The application server cluster and the data server cluster are in communication connection through a gigabit local-area network. The platform has the advantages that any quantity of loads (common/ gateway) and load balances are supported so that expandability of a system is greatly strengthened; on-line-user supporting capacity of the system is greatly strengthened through service resources provided by a large quantity of nodes; system operation and maintenance cost is reduced by not limiting start sequence of the nodes; through dynamic cluster entrance and exit of the nodes and automatic reconstruction of a system cluster network, a system dynamic expansion capability is increased, system stop time is reduced and system operation and maintenance cost is reduced; and through distributed storage of data, supporting capability for large-capacity data and system-fault resistance capability of the system are improved.
Description
Technical field
The present invention relates to a kind of real-time messages switching plane, relate in particular to a kind of real-time messages switching plane and distributed type assemblies construction method.
Background technology
Real-time messages switching plane (Server) has been expanded standard x MPP agreement, realize group, department, meeting, file transfer and share, the functions such as voice, video, that database adopts is mnesia, between Servers-all, copies and the distributed storage data.Limit by initial designs, there is following problem in original system (ejabberd2111):
(1) group system extended capability limited (10~20 of the upper limits)
Original system only has the server node of a type, and every station server node is all for external client provides service, and internodal interconnection and exchanges data are undertaken by the long latticed cluster network be connected to form of internodal TCP.In the network of N node, only internodal connection just needs N* (N-1)/2, supposes the cluster network of 20 nodes, and between node, data communication just will consume 190 connections, and 100 nodes will reach 4950 connections.In the situation that number of servers increases, if the more fearful clustered node transformation of the internal consumption of whole system resource is estimated as 20, support 5000~10000 calculating of online user by each node, the server cluster system of 20 nodes estimates that supporting online number of users is 100000~200000.
(2) database (mnesia) enabling capabilities is limited
Original system adopts mnesia database storage user data, comprises account, deferred information, session information, user's business card etc.By the Distributed Storage mechanism of mnesia, user profile is stored on Servers-all, and each node needs to store all or part of data, and data storage each time all will complete one time db transaction on all nodes.Surpass at 10~20 o'clock at number of nodes, between node, data synchronously will consume the ample resources of distributed cluster system.
Summary of the invention
Purpose of the present invention provides a kind of real-time messages switching plane and distributed type assemblies construction method with regard to being in order to address the above problem.
The present invention is achieved through the following technical solutions above-mentioned purpose:
The present invention includes application server cluster and data server cluster, described application server cluster is connected by gigabit LAN ruton letter with described data server cluster.
Particularly, described data server cluster comprises data server, management server, the first sql server, the second sql server and virtual server, described management server respectively with described data server and the first sql server and described the second sql server communication connection, described virtual server respectively with described the first sql server and described the second sql server communication connection.
Described data server comprises a plurality of, and each described data server is connected with described management server communication respectively.
Particularly, described application server comprises chat server, File Agent server and voice and video server, described virtual server and the communication connection of described chat server, described virtual server is monitored described File Agent server and described voice and video server.
Real-time messages switching plane distributed type assemblies construction method of the present invention:
Step 1: the server in described application server cluster is divided into two types of ordinary node and gateway nodes;
(1) ordinary node: provide access service to external client, complete the server of the practical business functions such as login, checking, session management, information processing.
(2) gateway node: fully transparent to external client, directly do not connect external client, only with each ordinary node, be connected, possess monitoring ordinary node survival condition, the function such as forwarding messages between ordinary node.
Step 2: all nodal informations are stored in a global node table, by system manager, node oneself, jointly safeguard the relevant information in the global node table; Each node is increased a node data after installing in the global node table by the system manager, the nodename, node type that nodename appointment when installing is set for this node carry on a shoulder pole thousand character types (0/1), whether enable be set to 1/ enable, whether move should be set to 0/ stop, ip is real ip address, describe can not fill out, safeguarded while by node, controlling oneself startup start-up time; If in the global node table, do not record nodal information, this node can't add cluster, and can't communication between other node.
(1) node starts: node should be guaranteed the correct of information in the global node table before startup, and whether enables (isenabled) and be set to 1/ and enable; During startup, the running status (isstart) that at first the node service upgrades current node is that 1/ startup, start-up time (start_time) are the current time, then search all nodes in running status in the global node table, and be cached in the node server internal memory; So far gateway node, ordinary node carry out next step action by different algorithms;
(2) node stop: node stop is divided into and normally stops and abending, no matter which kind of stops situation, monitor node all will be received node stop message, and when online without gateway node, ordinary node runs into the message that need to be forwarded to other node, can be connected directly to this node, and monitor the presence of this node, when monitored node stops, monitor node can be received node stop message equally, it is deleted from local line node buffer memory, and be synchronized to the global node table.
(3) cluster network builds automatically: the application service cluster consists of several nodes, wherein gateway node forms the ground floor loop network of cluster, not monitoring mutually between gateway node, ordinary node forms the second layer network of cluster, monitoring mutually between ordinary node (without gateway node time except);
Step 3: overall session management, after the client login, can record session information at connected node, to keep session status.After establishment distributed type assemblies network, session information should be shared in all nodes, therefore creates overall session table, by its same maintenance of all nodes.
Step 4: load balancing, chat server cluster load balancing is divided into the load balancing of ordinary node and the load balancing of gateway node.
(1) ordinary node load balancing, ordinary node curstomer-oriented end provides access and Message Processing service, cluster network comprise several ordinary nodes but externally domain name or IP only have one, utilize its IP tunnel or IP route technology to realize load balancing.
(2) gateway node load balancing, in the group system of a plurality of gateway nodes is arranged, gateway load-balancing is moved according to the following steps:
When 1) each ordinary node needs forwarding information, first in local cache, search online gateway node;
2) select a gateway node by the load-balancing algorithm of setting;
3) destination node and forwarding information group bag are sent to gateway node;
4) after gateway node is received a forwarding bag, judge that whether destination node is online;
5) if online, forwarding information;
6) if not online, proceed to information off-line and process.
Step 5: the message route, the message in Server is mainly and is divided into tri-kinds of presence, iq, message.
(1) presence, in group system, sends a presence information and follows following routing rule:
Whether 1) search targeted customer (xxx@yyycom/zzz) logins;
2) targeted customer does not login and abandons this message;
3) targeted customer logins, judges whether this node of its login node;
4) be this node, directly send message;
5) not this node, by destination node and message groups bag, by load-balancing algorithm, select a gateway node, forwarding messages;
6) after ordinary node receives that a gateway node is sent message, judge destination node whether oneself;
7) if not, abandon this message;
8) if judge whether the targeted customer logins at this node;
9) if not, abandon this message;
10) if to the targeted customer, send this message.
(2) iq, in group system, sends an iq information and follows following routing rule:
1) search iq message and whether comprise the targeted customer;
2) if nothing, directly this node is processed;
3), if having, search targeted customer (xxx@yyycom/zzz) and whether login;
4) targeted customer does not login and returns to error message;
5) targeted customer logins, judges whether this node of its login node;
6) be this node, directly send message;
7) not this node, by destination node and message groups bag, by load-balancing algorithm, select a gateway node, forwarding messages;
8) after ordinary node receives that a gateway node is sent message, judge destination node whether oneself;
9) if not, abandon this message;
10) if judge whether the targeted customer logins at this node;
11) if not, abandon this message;
12) if to the targeted customer, send this message.
(3) message, in group system, sends a message information and follows following routing rule:
Whether the targeted customer who 1) searches message message specifies specific resources (res);
2) if do not specify, as: xxx@yyycom searches the session that this User Priority is the highest in overall session;
3), if specify, judge whether this user logins;
4), if the targeted customer does not login, proceed to off-line message Message Processing;
5), if the targeted customer logins, judge whether this node of its login node;
6) be this node, directly send message;
7) not this node, by destination node and message groups bag, by load-balancing algorithm, select a gateway node, forwarding messages;
8) after ordinary node receives that a gateway node is sent message, judge destination node whether oneself;
9) if not, abandon this message;
10) if judge whether the targeted customer logins at this node;
11) if not, proceed to deferred information and process (may disconnect and causing because client is abnormal);
12) if to the targeted customer, send this message.
Step 6: File Agent server cluster, Server is when the File Agent server cluster, consideration is mapped to the agency service of each node the particular port of external domain name or IP, be that client is distributed a proxy service node by server before carrying out the File Agent transmission, guarantee that the both sides that participate in transmission are connected to same proxy service node.
Step 7: voice and video server cluster, Server is when the voice and video server cluster, consideration is mapped to the service of each node the particular port of external domain name or IP, be that client is distributed a service node by server before carrying out Stream Data Transmission, guarantee that the client in many ways that participates in transmission all is connected to same service node.
Step 8: database design adopts the my sql server to carry out distributed type assemblies, by 1 management node, 2 sql server nodes, 2 data nodes, formed, data fragmentation is stored, is copied 2 parts, be stored in respectively in each back end, whole at a complete database of logic formation, externally by virtual server, by sql server, provide data access service; Later stage increases the quantity of sql server node and back end according to practical operation situation such as the growth of data volume, visit capacity increases
Further, start and comprise and gateway node and ordinary node carry out next step action by different algorithms according to the described node of the step in step 2 (1):
1) gateway node: gateway node is searched the ordinary node of all operations in the global node table, notifies each common joint oneself to reach the standard grade, and initiates the monitoring to all ordinary nodes; After gateway node is received the on-line message of ordinary node, it is updated in local line node buffer memory.
2) ordinary node: ordinary node is searched the gateway node of all operations in the global node table, notify each gateway node oneself to reach the standard grade, and initiate the monitoring to all gateway nodes, after ordinary node is received the on-line message of gateway node, it is updated in local line node buffer memory.
Further, according to the described node stop of the step in step 2 (2), comprise and normally stop and abending,
1) gateway node receive ordinary node stop message after, it is deleted from local line node buffer memory, and is synchronized to the global node table:
2) ordinary node receive gateway node stop message after, it is deleted from local line node buffer memory, and is synchronized to the global node table.
Beneficial effect of the present invention is:
A kind of real-time messages switching plane and the distributed type assemblies construction method of providing of the present invention has:
1, support any amount node (common/gateway) and load balancing, greatly strengthen the system expandability;
2, the Service Source provided by great deal of nodes, greatly strengthen online user's enabling capabilities of system;
3, reduce the system O&M cost by not limiting the node boot sequence;
4, by node dynamically add and break away from cluster, automatic reconfiguration system cluster network, increased system dynamic capacity-expanding ability, reduce the system dwell time, reduce the system O&M cost;
5, by the distributed row storage of data data, the increase system is to the enabling capabilities of Large Volume Data, to system failure resistivity.
The accompanying drawing explanation
Fig. 1 is the structured flowchart of data server cluster of the present invention.
Fig. 2 is the structured flowchart of application server cluster of the present invention.
Fig. 3 is that real-time messages switching plane node of the present invention starts schematic diagram.
Fig. 4 is the initial schematic diagram without the gateway cluster network struction of the present invention.
Fig. 5 is schematic diagram in the middle of the node without the gateway cluster network struction of the present invention builds.
Fig. 6 is that the node without the gateway cluster network struction of the present invention has built schematic diagram.
Fig. 7 is that the cluster network of real-time messages switching plane of the present invention and distributed type assemblies construction method builds schematic diagram.
Embodiment
Below in conjunction with accompanying drawing, the invention will be further described:
As Fig. 1, Fig. 2, Fig. 3, Fig. 4, Fig. 5, shown in Fig. 6 and Fig. 7: the present invention includes application server cluster and data server cluster, the data server cluster comprises a plurality of data servers, management server, the first sql server, the second sql server, virtual server, chat server, File Agent server and voice and video server, management server respectively with each data server and the first sql server and the communication connection of the second sql server, virtual server respectively with the first sql server, the second sql server is connected with application server communication, described virtual server is monitored described File Agent server and described voice and video server.
Real-time messages switching plane distributed type assemblies construction method of the present invention:
Step 1: the server in described application server cluster is divided into two types of ordinary node and gateway nodes;
(1) ordinary node: provide access service to external client, complete the server of the practical business functions such as login, checking, session management, information processing.
(2) gateway node: fully transparent to external client, directly do not connect external client, only with each ordinary node, be connected, possess monitoring ordinary node survival condition, the function such as forwarding messages between ordinary node.
Step 2: all nodal informations are stored in a global node table, by system manager, node oneself its with safeguarding the relevant information in the global node table; Each node is increased a node data after installing in the global node table by the system manager, arrange nodename appointment when installing nodename, character types (0/1) that node type is served as for this node, whether enable be set to 1/ enable, whether move should be set to 0/ stop, ip is real ip address, description can not filled out, safeguarded while being initiated self by node start-up time; If in the global node table, do not record nodal information, this node can't add cluster, and can't communication between other node.
(1) node starts: node should be guaranteed the correct of information in the global node table before startup, and whether enables (isenabled) and be set to 1/ and enable; During startup, the running status (isstart) that at first the node service upgrades current node is that 1/ startup, start-up time (start_time) are the current time, then search all nodes in running status in the global node table, and be cached in the node server internal memory; So far gateway node, ordinary node carry out next step action by different algorithms;
1) gateway node: gateway node is searched the ordinary node of all operations in the global node table, notify each ordinary node oneself reach the standard grade, and initiate the monitoring to all ordinary nodes; After gateway node is received the on-line message of common joint mid point, it is updated in local line node buffer memory.
2) ordinary node: ordinary node is searched the gateway node of all operations in the global node table, notify each gateway node oneself to reach the standard grade, and initiate the monitoring to all gateway nodes, ordinary node is updated to it in local line node buffer memory after receiving the on-line message of gateway node.
(2) node stop: node stop is divided into and normally stops and abending, no matter which kind of stops situation, monitor node all will be received node stop message, and when online without gateway node, ordinary node runs into the message that need to be forwarded to other node, can be connected directly to this node, and monitor the presence of this node, when monitored node stops, monitor node can be received node stop message equally, it is deleted from local line node buffer memory, and be synchronized to the global node table.
1) gateway node receive ordinary node stop message after, it is deleted from local line node buffer memory, and is synchronized to the global node table;
2) ordinary node receive gateway node stop message after, it is deleted from local line node buffer memory, and is synchronized to the global node table.
(3) cluster network builds automatically: the application service cluster consists of several nodes, wherein gateway node forms the ground floor loop network of cluster, monitoring mutually between gateway node, ordinary node forms monitoring mutually between the second layer network ordinary node of cluster (without gateway node time except);
Step 3: overall session management, after the client login, can record session information at connected node, to keep session status.After setting up the distributed type assemblies network, session information should be shared in all nodes, therefore creates overall session table, by all nodes, is jointly safeguarded.
Step 4: load balancing, chat server cluster load balancing is divided into the load balancing of ordinary node and the load balancing of gateway node.
(1) ordinary node load balancing, ordinary node curstomer-oriented end provides access and Message Processing service, and a cluster network comprises several ordinary nodes, but external domain name or IP only have one, utilize its IP tunnel or IP route technology to realize load balancing.
(2) gateway node load balancing, in the group system of a plurality of gateway nodes is arranged, gateway load-balancing is moved according to the following steps:
When 1) each ordinary node needs forwarding information, first in local cache, search online gateway node;
2) select a gateway node by the load-balancing algorithm of setting;
3) destination node and forwarding information group bag are sent to gateway node;
4) after gateway node is received a forwarding bag, judge that whether destination node is online;
5) if online, forwarding information;
6) if not online, proceed to information off-line and process.
Step 5: the message route, the message in Server is mainly and is divided into tri-kinds of presenceiq, message.
(1) presence, in group system, sends a presence information and follows following routing rule:
Whether 1) search targeted customer (xxx@yyycom/zzz) logins;
2) targeted customer does not login and abandons this message;
3) targeted customer logins, judges whether this node of its login node;
4) be this node, directly send message;
5) not this node, by destination node and message groups bag, by load-balancing algorithm, select a gateway node, forwarding messages;
6) after ordinary node receives that a gateway node is sent message, judge destination node whether oneself;
7) if not, abandon this message;
8) if judge whether the targeted customer logins at this node;
9) if not, abandon this message;
10) if to the targeted customer, send this message.
(2) iq, in group system, sends an iq information and follows following routing rule:
1) search iq message and whether comprise the targeted customer;
2) if nothing, directly this node is processed;
3), if having, search targeted customer (xxx@yyycom/zzz) and whether login;
4) targeted customer does not login and returns to error message;
5) targeted customer logins, judges whether this node of its login node;
6) be that this node directly sends message;
7) not this node, by destination node and message groups bag, by load-balancing algorithm, select a gateway node, forwarding messages;
8) after ordinary node receives that a gateway node is sent message, judge destination node whether oneself;
9) if not, abandon this message;
10) if judge whether the targeted customer logins at this node;
11) if not, abandon this message;
12) if to the targeted customer, send this message.
(3)message
In group system, send a message information and follow following routing rule:
Whether the targeted customer who 1) searches message message specifies specific resources (res);
2) if do not specify, as: xxx@yyycom searches the session that this User Priority is the highest in overall session;
3), if specify, judge whether this user logins;
4), if the targeted customer does not login, proceed to off-line message Message Processing;
5), if the targeted customer logins, judge whether this node of its login node;
6) be this node, directly send message;
7) not this node, by destination node and message groups bag, by load-balancing algorithm, select a gateway node, forwarding messages;
8) after ordinary node receives that a gateway node is sent message, judge destination node whether oneself;
9) if not, abandon this message;
10) if judge whether the targeted customer logins at this node;
11) if not, proceed to deferred information and process (may disconnect and causing because client is abnormal);
12) if to the targeted customer, send this message.
Step 6: File Agent server cluster, Server is when the File Agent server cluster, consideration is mapped to the agency service of each node the particular port of external domain name or IP, be that client is distributed a proxy service node by server before carrying out the File Agent transmission, guarantee that the both sides that participate in transmission are connected to same proxy service node.
Step 7: voice and video server cluster, Server is when the voice and video server cluster, consideration is mapped to the service of each node the particular port of external domain name or IP, be that client is distributed a service node by server before carrying out Stream Data Transmission, guarantee that the client in many ways that participates in transmission all is connected to same service node.
Step 8: database design adopts the my sql server to carry out distributed type assemblies, by 1 management node, 2 sql server nodes, 2 data nodes, formed, 2 parts of data fragmentation storage replications, be stored in respectively in each back end, whole at a complete database of logic formation, externally by virtual server, by sql server, provide data access service; Later stage increases the quantity of sql server node and back end according to the practical operation situation such as growth visit capacity increase of data volume.
As shown in Figure 3: wherein G is gateway node, N is ordinary node, and G1 is that ordinary node, the G3 that searches operation is to be that ordinary node, the N3 that searches operation is the notice ordinary node and faces control with new running status/start-up time, N2 for notifying ordinary node monitoring, N1 for following new running status/start-up time, G2.
As shown in Figure 4, Figure 5 and Figure 6: when the needs starter node, ordinary node is connected, and finally completes ordinary node and starts.
As shown in Figure 7: when ordinary node need to start, the gateway node 1 that ordinary node 2 is selected is started, the gateway node 1 that ordinary node 3 is selected is started, the gateway node 1 that ordinary node 4 is selected is started, the ordinary node 2,3,4 that gateway node 5 is selected is started, and the gateway node 1 and 5 that ordinary node 6 is selected is started.
Claims (7)
1. a real-time messages switching plane, it is characterized in that: comprise application server cluster and data server cluster, described application server cluster is connected by gigabit LAN ruton letter with described data server cluster.
2. real-time messages switching plane according to claim 1, it is characterized in that: described data server cluster comprises data server, management server, the first sql server, the second sql server and virtual server, described management server respectively with described data server and the first sql server and described the second sql server communication connection, described virtual server respectively with described the first sql server and described the second sql server communication connection.
3. real-time messages switching plane according to claim 2, it is characterized in that: described data server comprises a plurality of, each described data server is connected with described management server communication respectively.
4. real-time messages switching plane according to claim 1 and 2, it is characterized in that: described application server comprises chat server, File Agent server and voice and video server, described virtual server and the communication connection of described chat server, described virtual server is monitored described File Agent server and described voice and video server.
5. a real-time messages switching plane distributed type assemblies construction method is characterized in that:
Step 1: the server in described application server cluster is divided into two types of ordinary node and gateway nodes;
(1) ordinary node: provide access service to external client, complete the server of the practical business functions such as login, checking, session management, information processing;
(2) gateway node: fully transparent to external client, directly do not connect external client, only with each ordinary node, be connected, possess monitoring ordinary node survival condition, the function such as forwarding messages between ordinary node.
Step 2: all nodal informations are stored in a global node table, by system manager, node oneself, jointly safeguard the relevant information in the global node table; Each node is increased a node data after installing in the global node table by the system manager, the nodename, node type that nodename appointment when installing is set for this node carry on a shoulder pole thousand character types (0/1), whether enable be set to 1/ enable, whether move should be set to 0/ stop, ip is real ip address, describe can not fill out, safeguarded while by node, controlling oneself startup start-up time; If in the global node table, do not record nodal information, this node can't add cluster, and can't communication between other node.
(1) node starts: node should be guaranteed the correct of information in the global node table before startup, and whether enables (isenabled) and be set to 1/ and enable; During startup, the running status (isstart) that at first the node service upgrades current node is that 1/ startup, start-up time (start_time) are the current time, then search all nodes in running status in the global node table, and be cached in the node server internal memory; So far gateway node, ordinary node carry out next step action by different algorithms;
(2) node stop: node stop is divided into and normally stops and abending, no matter which kind of stops situation, monitor node all will be received node stop message, and when online without gateway node, ordinary node runs into the message that need to be forwarded to other node, can be connected directly to this node, and monitor the presence of this node, when monitored node stops, monitor node can be received node stop message equally, it is deleted from local line node buffer memory, and be synchronized to the global node table;
(3) cluster network builds automatically: the application service cluster consists of several nodes, wherein gateway node forms the ground floor loop network of cluster, not monitoring mutually between gateway node, ordinary node forms the second layer network of cluster, monitoring mutually between ordinary node (without gateway node time except).
Step 3: overall session management, after the client login, can record session information at connected node, to keep session status.After setting up the distributed type assemblies network, session information should be prosperous altogether in all nodes, therefore creates overall session table, by all nodes, jointly safeguarded.
Step 4: load balancing, chat server cluster load balancing is divided into the load balancing of ordinary node and the load balancing of gateway node.
(1) ordinary node load balancing, ordinary node curstomer-oriented end provides access and Message Processing service, and a cluster network comprises several ordinary nodes, but external domain name or IP only have one, utilize its IP tunnel or IP route technology to realize load balancing;
(2) gateway node load balancing, in the group system of a plurality of gateway nodes is arranged, gateway load-balancing is moved according to the following steps:
When 1) each ordinary node needs forwarding information, first in local cache, search online gateway node;
2) select a gateway node by the load-balancing algorithm of setting;
3) destination node and forwarding information group bag are sent to gateway node;
4) after gateway node is received a forwarding bag, judge that whether destination node is online;
5) if online, forwarding information;
6) if not online, proceed to information off-line and process.
Step 5: the message route, the message in Server is mainly and is divided into tri-kinds of presence, iq, message.
(1) presence, in group system, sends a presence information and follows following routing rule:
Whether 1) search targeted customer (xxx@yyycom/zzz) logins;
2) targeted customer does not login and abandons this message;
3) targeted customer logins, judges whether this node of its login node;
4) be this node, directly send message;
5) not this node, by destination node and message groups bag, by load-balancing algorithm, select a gateway node, forwarding messages;
6) after ordinary node receives that a gateway node is sent message, judge destination node whether oneself;
7) if not, abandon this message;
8) if judge whether the targeted customer logins at this node;
9) if not, abandon this message;
10) if to the targeted customer, send this message.
(2) iq, in group system, sends an iq information and follows following routing rule:
1) search iq message and whether comprise the targeted customer;
2) if nothing, directly this node is processed;
3), if having, search targeted customer (xxx@yyycom/zzz) and whether login;
4) targeted customer does not login and returns to error message;
5) targeted customer logins, judges whether this node of its login node;
6) be this node, directly send message;
7) not this node, by destination node and message groups bag, by load-balancing algorithm, select a gateway node, forwarding messages;
8) after ordinary node receives that a gateway node is sent message, judge destination node whether oneself;
9) if not, abandon this message;
10) if judge whether the targeted customer logins at this node;
11) if not, abandon this message;
12) if to the targeted customer, send this message.
(3) message, in group system, sends a message information and follows following routing rule:
Whether the targeted customer who 1) searches message message specifies specific resources (res);
2) if do not specify, as: xxx@yyycom searches the session that this User Priority is the highest in overall session;
3), if specify, judge whether this user logins;
4), if the targeted customer does not login, proceed to off-line message Message Processing;
5), if the targeted customer logins, judge whether this node of its login node;
6) be this node, directly send message;
7) not this node, by destination node and message groups bag, by load-balancing algorithm, select a gateway node, forwarding messages;
8) after ordinary node receives that a gateway node is sent message, judge destination node whether oneself;
9) if not, abandon this message;
10) if judge whether the targeted customer logins at this node;
11) if not, proceed to deferred information and process (may disconnect and causing because client is abnormal);
12) if to the targeted customer, send this message.
Step 6: File Agent server cluster, Server is when the File Agent server cluster, consideration is mapped to the agency service of each node the particular port of external domain name or IP, be that client is distributed a proxy service node by server before carrying out the File Agent transmission, guarantee that the both sides that participate in transmission are connected to same proxy service node.
Step 7: voice and video server cluster, Server is when the voice and video server cluster, consideration is mapped to the service of each node the particular port of external domain name or IP, be that client is distributed a service node by server before carrying out Stream Data Transmission, guarantee that the client in many ways that participates in transmission all is connected to same service node.
Step 8: database design adopts the my sql server to carry out distributed type assemblies, by 1 management node, 2 sql server nodes, 2 data nodes, formed, data fragmentation is stored, is copied 2 parts, be stored in respectively in each back end, whole at a complete database of logic formation, externally by virtual server, by sql server, provide data access service; Later stage increases the quantity of sql server node and back end according to practical operation situation such as the growth of data volume, visit capacity increases.
6. real-time messages switching plane distributed type assemblies construction method according to claim 5 is characterized in that: according to the described node of the step in step 2 (1), start and comprise and gateway node and ordinary node carry out next step action by different algorithms:
1) gateway node: gateway node is searched the ordinary node of all operations in the global node table, notifies each ordinary node oneself to reach the standard grade, and initiates the monitoring to all ordinary nodes; After gateway node is received the on-line message of ordinary node, it is updated in local line node buffer memory;
2) ordinary node: ordinary node is searched the gateway node of all operations in the global node table and is notified each gateway node oneself to reach the standard grade, and initiation is to the monitoring of all gateway nodes, after ordinary node is received the on-line message of gateway node, it is updated in local line node buffer memory.
7. real-time messages switching plane distributed type assemblies construction method according to claim 5 is characterized in that: comprises and normally stops and abending according to the described node stop of the step in step 2 (2),
1) gateway node receive ordinary node stop message after, it is deleted from local line node buffer memory, and is synchronized to the global node table:
2) ordinary node receive gateway node stop message after, it is deleted from local line node buffer memory, and is synchronized to the global node table.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2013102877013A CN103475566A (en) | 2013-07-10 | 2013-07-10 | Real-time message exchange platform and distributed cluster establishment method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2013102877013A CN103475566A (en) | 2013-07-10 | 2013-07-10 | Real-time message exchange platform and distributed cluster establishment method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103475566A true CN103475566A (en) | 2013-12-25 |
Family
ID=49800273
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2013102877013A Pending CN103475566A (en) | 2013-07-10 | 2013-07-10 | Real-time message exchange platform and distributed cluster establishment method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103475566A (en) |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104994168A (en) * | 2015-07-14 | 2015-10-21 | 苏州科达科技股份有限公司 | distributed storage method and distributed storage system |
WO2016062079A1 (en) * | 2014-10-21 | 2016-04-28 | 中兴通讯股份有限公司 | Off-line message processing method and apparatus |
WO2017028697A1 (en) * | 2015-08-17 | 2017-02-23 | 阿里巴巴集团控股有限公司 | Method and device for growing or shrinking computer cluster |
CN106570074A (en) * | 2016-10-14 | 2017-04-19 | 深圳前海微众银行股份有限公司 | Distributed database system and implementation method thereof |
CN106844399A (en) * | 2015-12-07 | 2017-06-13 | 中兴通讯股份有限公司 | Distributed data base system and its adaptive approach |
CN109167719A (en) * | 2018-08-16 | 2019-01-08 | 广州爽游网络科技有限公司 | A kind of super large community implementation method with content intelligence isolation features |
CN109639773A (en) * | 2018-11-26 | 2019-04-16 | 中国船舶重工集团公司第七六研究所 | A kind of the distributed data cluster control system and its method of dynamic construction |
CN109684081A (en) * | 2018-12-11 | 2019-04-26 | 北京数盾信息科技有限公司 | A kind of allocation processing method of load balancing in cluster |
CN109792406A (en) * | 2018-07-27 | 2019-05-21 | 袁振南 | Message delivery method, device and storage medium in server cluster |
CN110334543A (en) * | 2019-06-20 | 2019-10-15 | 广东技术师范大学天河学院 | A kind of block chain knowledge system and its application method based on calculation power |
CN110502266A (en) * | 2019-08-26 | 2019-11-26 | 中电九天智能科技有限公司 | A kind of seamless update method of CIM server cluster |
CN111817967A (en) * | 2020-08-28 | 2020-10-23 | 支付宝(杭州)信息技术有限公司 | Communication optimization system, registration method and message forwarding method of block chain network |
CN112488800A (en) * | 2020-12-15 | 2021-03-12 | 北京易兴元石化科技有限公司 | Network-based petroleum operation data processing method and device and storage medium |
CN112491747A (en) * | 2020-11-25 | 2021-03-12 | 湖南麒麟信安科技股份有限公司 | Application method of high-availability data exchange cluster |
CN112511628A (en) * | 2020-11-30 | 2021-03-16 | 银盛支付服务股份有限公司 | Distributed real-time message pushing method |
CN113676477A (en) * | 2021-08-20 | 2021-11-19 | 中邮科通信技术股份有限公司 | Method for realizing Socket session cluster communication based on star network and memory routing table |
CN113986522A (en) * | 2021-08-29 | 2022-01-28 | 中盾创新数字科技(北京)有限公司 | Load balancing-based distributed storage server capacity expansion system |
CN114205889A (en) * | 2021-12-09 | 2022-03-18 | 中山大学 | Zenoh-based communication method in cross-local-area-network distributed system |
CN115988005A (en) * | 2022-11-08 | 2023-04-18 | 北京太格时代自动化系统设备有限公司 | Method for signal synchronization by adopting distributed cluster |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101242375A (en) * | 2008-03-12 | 2008-08-13 | 程清(上海)信息科技有限公司 | An Internet cross-website and cross-domain instant messaging system |
CN102082826A (en) * | 2010-12-30 | 2011-06-01 | 上海华冠电子设备有限责任公司 | Electric energy data cloud collection system and method |
CN102546807A (en) * | 2012-01-17 | 2012-07-04 | 上海宏舜电子科技有限公司 | Network service system based on cloud computing architecture |
CN102546646A (en) * | 2012-01-17 | 2012-07-04 | 深圳市乐唯科技开发有限公司 | System and method for realizing voice talkback function |
US20120226797A1 (en) * | 2011-03-01 | 2012-09-06 | Cisco Technology, Inc. | Active Load Distribution for Control Plane Traffic Using a Messaging and Presence Protocol |
CN103049854A (en) * | 2012-12-20 | 2013-04-17 | 雷华 | 3D (three dimensional) network shopping platform |
-
2013
- 2013-07-10 CN CN2013102877013A patent/CN103475566A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101242375A (en) * | 2008-03-12 | 2008-08-13 | 程清(上海)信息科技有限公司 | An Internet cross-website and cross-domain instant messaging system |
CN102082826A (en) * | 2010-12-30 | 2011-06-01 | 上海华冠电子设备有限责任公司 | Electric energy data cloud collection system and method |
US20120226797A1 (en) * | 2011-03-01 | 2012-09-06 | Cisco Technology, Inc. | Active Load Distribution for Control Plane Traffic Using a Messaging and Presence Protocol |
CN102546807A (en) * | 2012-01-17 | 2012-07-04 | 上海宏舜电子科技有限公司 | Network service system based on cloud computing architecture |
CN102546646A (en) * | 2012-01-17 | 2012-07-04 | 深圳市乐唯科技开发有限公司 | System and method for realizing voice talkback function |
CN103049854A (en) * | 2012-12-20 | 2013-04-17 | 雷华 | 3D (three dimensional) network shopping platform |
Non-Patent Citations (1)
Title |
---|
倪庆峰: "基于集群技术的web即时通信系统", 《中国优秀硕士学位论文全文数据库 信息科技辑》, 15 May 2011 (2011-05-15) * |
Cited By (31)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016062079A1 (en) * | 2014-10-21 | 2016-04-28 | 中兴通讯股份有限公司 | Off-line message processing method and apparatus |
CN104994168B (en) * | 2015-07-14 | 2018-05-01 | 苏州科达科技股份有限公司 | Distributed storage method and distributed memory system |
CN104994168A (en) * | 2015-07-14 | 2015-10-21 | 苏州科达科技股份有限公司 | distributed storage method and distributed storage system |
WO2017028697A1 (en) * | 2015-08-17 | 2017-02-23 | 阿里巴巴集团控股有限公司 | Method and device for growing or shrinking computer cluster |
CN106844399B (en) * | 2015-12-07 | 2022-08-09 | 中兴通讯股份有限公司 | Distributed database system and self-adaptive method thereof |
CN106844399A (en) * | 2015-12-07 | 2017-06-13 | 中兴通讯股份有限公司 | Distributed data base system and its adaptive approach |
CN106570074A (en) * | 2016-10-14 | 2017-04-19 | 深圳前海微众银行股份有限公司 | Distributed database system and implementation method thereof |
CN106570074B (en) * | 2016-10-14 | 2020-07-10 | 深圳前海微众银行股份有限公司 | Distributed database system and implementation method thereof |
CN109792406A (en) * | 2018-07-27 | 2019-05-21 | 袁振南 | Message delivery method, device and storage medium in server cluster |
CN109792406B (en) * | 2018-07-27 | 2021-06-18 | 袁振南 | Message transmission method, device and storage medium in server cluster |
CN109167719A (en) * | 2018-08-16 | 2019-01-08 | 广州爽游网络科技有限公司 | A kind of super large community implementation method with content intelligence isolation features |
CN109639773B (en) * | 2018-11-26 | 2021-04-13 | 中国船舶重工集团公司第七一六研究所 | Dynamically constructed distributed data cluster control system and method thereof |
CN109639773A (en) * | 2018-11-26 | 2019-04-16 | 中国船舶重工集团公司第七六研究所 | A kind of the distributed data cluster control system and its method of dynamic construction |
CN109684081A (en) * | 2018-12-11 | 2019-04-26 | 北京数盾信息科技有限公司 | A kind of allocation processing method of load balancing in cluster |
CN109684081B (en) * | 2018-12-11 | 2023-11-07 | 北京数盾信息科技有限公司 | Distribution processing method for load balancing in cluster |
CN110334543A (en) * | 2019-06-20 | 2019-10-15 | 广东技术师范大学天河学院 | A kind of block chain knowledge system and its application method based on calculation power |
CN110334543B (en) * | 2019-06-20 | 2023-01-17 | 广东技术师范大学天河学院 | Calculation power-based block chain knowledge system and use method thereof |
CN110502266A (en) * | 2019-08-26 | 2019-11-26 | 中电九天智能科技有限公司 | A kind of seamless update method of CIM server cluster |
CN111817967A (en) * | 2020-08-28 | 2020-10-23 | 支付宝(杭州)信息技术有限公司 | Communication optimization system, registration method and message forwarding method of block chain network |
US11388017B2 (en) | 2020-08-28 | 2022-07-12 | Alipay (Hangzhou) Information Technology Co., Ltd. | Communication optimization systems of blockchain network, registration methods and message forwarding methods |
CN112491747A (en) * | 2020-11-25 | 2021-03-12 | 湖南麒麟信安科技股份有限公司 | Application method of high-availability data exchange cluster |
CN112491747B (en) * | 2020-11-25 | 2022-06-14 | 湖南麒麟信安科技股份有限公司 | Application method of high-availability data exchange cluster |
CN112511628A (en) * | 2020-11-30 | 2021-03-16 | 银盛支付服务股份有限公司 | Distributed real-time message pushing method |
CN112488800A (en) * | 2020-12-15 | 2021-03-12 | 北京易兴元石化科技有限公司 | Network-based petroleum operation data processing method and device and storage medium |
CN112488800B (en) * | 2020-12-15 | 2024-01-23 | 北京易兴元石化科技有限公司 | Petroleum operation data processing method, device and storage medium based on network |
CN113676477A (en) * | 2021-08-20 | 2021-11-19 | 中邮科通信技术股份有限公司 | Method for realizing Socket session cluster communication based on star network and memory routing table |
CN113986522A (en) * | 2021-08-29 | 2022-01-28 | 中盾创新数字科技(北京)有限公司 | Load balancing-based distributed storage server capacity expansion system |
CN114205889A (en) * | 2021-12-09 | 2022-03-18 | 中山大学 | Zenoh-based communication method in cross-local-area-network distributed system |
CN114205889B (en) * | 2021-12-09 | 2023-10-20 | 中山大学 | Zenoh-based inter-LAN distributed system communication method |
CN115988005A (en) * | 2022-11-08 | 2023-04-18 | 北京太格时代自动化系统设备有限公司 | Method for signal synchronization by adopting distributed cluster |
CN115988005B (en) * | 2022-11-08 | 2023-07-18 | 北京太格时代电气股份有限公司 | Method for synchronizing signals by adopting distributed clusters |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103475566A (en) | Real-time message exchange platform and distributed cluster establishment method | |
US8432791B1 (en) | Network route injections | |
US20130254415A1 (en) | Routing requests over a network | |
CN110351246A (en) | Server cluster system Socket management method and device | |
CN102045270B (en) | Chat server system, method for constructing chat server system and chat system | |
CN103581276B (en) | Cluster management device, system, service customer end and correlation method | |
JP4588704B2 (en) | Self-management mediation information flow | |
CN111277629A (en) | High-availability-based web high-concurrency system and method | |
CN103546572B (en) | A kind of cloudy storing networking system and method | |
CN104272708A (en) | Two level packet distribution with stateless first level packet distribution to a group of servers and stateful second level packet distribution to a server within the group | |
CN103747091A (en) | Distributed data synchronization system and method for embedded equipment | |
CN101014046B (en) | Method for integrating service location with service quality routing in service loading network | |
CN104380289B (en) | Service-aware distributed hash table is route | |
US20220131935A1 (en) | Service Unit Switching Method, System, and Device | |
CN111638957A (en) | Method for realizing cluster sharing type public cloud load balance | |
CN107689878A (en) | TCP length connection SiteServer LBSs based on name scheduling | |
CN114301823B (en) | Route notification method, device, equipment and storage medium | |
CN100505662C (en) | Method for setting point-to-point data channel | |
JP2007524325A (en) | Non-stop service system using voting and information updating and providing method in the system | |
CN114025002A (en) | MQTT information transmission method, system and communication equipment | |
CN103401951B (en) | Based on the elastic cloud distribution method of peer-to-peer architecture | |
CN111756780B (en) | Method for synchronizing connection information and load balancing system | |
CN108259205B (en) | Route publishing method and network equipment | |
CN104836734A (en) | Brand new instant messaging routing method and router | |
Jingjing et al. | The deployment of routing protocols in distributed control plane of SDN |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20131225 |