CN102833333A - Scalable network server group frame system - Google Patents
Scalable network server group frame system Download PDFInfo
- Publication number
- CN102833333A CN102833333A CN2012103030885A CN201210303088A CN102833333A CN 102833333 A CN102833333 A CN 102833333A CN 2012103030885 A CN2012103030885 A CN 2012103030885A CN 201210303088 A CN201210303088 A CN 201210303088A CN 102833333 A CN102833333 A CN 102833333A
- Authority
- CN
- China
- Prior art keywords
- server
- management
- management server
- service
- division
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Abstract
The invention discloses a scalable network server group frame system, which comprises a plurality of management servers and business servers. The scalable network server group frame system is of a tree-form structure; the management servers are connected with a plurality of management servers or a plurality of business servers; the management servers are positioned at root nodes and intermediate nodes of the tree-form structure, and the business servers are positioned at leaf nodes of the tree-form structure; each management server comprises a member management module and a business assistive processing module; and each business server comprises a business processing module and a member management module. According to the scalable network server group frame system disclosed by the invention, a business operation state can be monitored and conveniently checked and can be further managed efficiently; in addition, the server group frame system is higher in scalability and frame balance; and meanwhile, load migration can be achieved between the adjacent servers as far as possible, and resource consumption caused by load migration can be reduced.
Description
Technical field
The present invention relates to network information technology field, particularly a kind of Scalable Network Services device gang fight construction system.
Background technology
Along with the development of Internet technology, the deployment of server is to realizing that operation system is more and more important efficiently.As far as the system that forms by the separate unit server; Because the restriction of aspects such as the central processing unit of separate unit server, input/output interface processing speed; The server farm architecture of being made up of multiple servers is the substitute is in the request that can't satisfy the data volume of quick growth and handle multiple business simultaneously.
In practical application, a lot of operation systems have very strong dynamic, only with peak period business need configuration server resource then be unfavorable for resource multiplex and energy-conserving and environment-protective.At present; The dynamic expansion server has more ripe scheme; But not only supported dynamic expansion but support dynamically reduction efficiently, the Managed Solution of server zone is also relatively deficienter easily; Therefore, it is significant that research and development can keep the deployment scheme and the management method of telescopic server zone of operation system stable operation.
Summary of the invention
In order to overcome the shortcoming and deficiency that prior art exists, the present invention provides a kind of Scalable Network Services device gang fight construction system.
The technical scheme that the present invention adopted:
A kind of Scalable Network Services device gang fight construction system comprises several management servers and service server, and said Scalable Network Services device gang fight structure is tree-like framework;
Said management server connects several management servers or several service servers; Said management server is positioned at the root node or the intermediate node position of tree-like framework, and service server is positioned at the leaf node position of tree-like framework;
Said management server comprises member management module and operation assistance processing module;
Member management module on the said management server that is positioned at root node is used for when load surpasses the division threshold value, sending the division request to the keeper; Be used to monitor the running status of subordinate's server,, then issue merge command to these servers if subordinate's server reaches the merging condition; Be used for after the division request that receives subordinate's server transmission, handling the division of subordinate's server; Be used for the division and the load reallocation of this machine of processing server; Be used to handle merging with subordinate's management server;
Member management module on the said management server that is positioned at intermediate node, be used to receive the division response that the upper management server issues after, carry out load reallocation and business migration; Be used for after receiving the merge command that the upper management server issues, carrying out load merging and business migration; Be used to monitor the running status of local server, state information be saved in database, and with the state information timed sending to higher level's management server; Be used for when load surpasses the division threshold value, sending the division request to higher level's management server; Be used to monitor the running status of subordinate's server,, then issue merge command to these servers if subordinate's server reaches the merging condition; Be used for after the division request that receives subordinate's server transmission, handling the division of subordinate's server; Be used for the nodes encoding management; Be used to refuse flag management;
Said operation assistance processing module, the service server under the user is searched in the operation assistance processing request that the operation assistance processing is asked or subordinate's management server is transmitted that is used for sending according to service server;
Comprise Service Processing Module and member management module on the said service server;
Member management module on the said service server, be used to receive the division response that the upper management server issues after, carry out load reallocation and business migration; Be used for after receiving the merge command that the upper management server issues, carrying out load merging and business migration; Be used to monitor the running status of local server, state information be saved in database, and with the state information timed sending to higher level's management server; Be used for when load surpasses the division threshold value, sending the division request to higher level's server; Be used for the nodes encoding management;
Said Service Processing Module is used for interface to the user being provided, and manages business according to service request from user.
Member management module in the said management server comprises monitoring thread, division thread and merges thread;
Member management module in the said service server comprises monitoring thread, division thread and merges thread;
Monitoring thread on the said management server that is positioned at root node is used to monitor the running status of local server and state information is saved in database; Be used to monitor the running status of subordinate's server, be used for sending the division request, send merge command to the subordinate's server that reaches the merging condition to the keeper; Be used for nodes encoding management and refusal flag management;
Monitoring thread on the said management server that is positioned at intermediate node is used to monitor the running status of local server, and state information is saved in database, and with the supreme one-level management server of state information timed sending; Be used to monitor the running status of subordinate's server, be used for sending the division request, send merge command to the subordinate's server that reaches the merging condition to higher level's server; Receive division response instruction and merging response instruction that the upper level server sends; Be used for nodes encoding management and refusal flag management;
Monitoring thread on the said service server is used to monitor the running status of local server, and state information is saved in database, and with the state information timed sending to higher level's management server; Be used for when load surpasses the division threshold value, upwards the one-level management server sends the division request; Be used to receive the division response instruction that the upper management server sends; Be used to receive the merge command that the upper management server sends; Be used for the nodes encoding management;
Division thread in said management server and the service server is used to handle load reallocation and business migration;
Merging thread in said management server and the service server is used to handle load merging and business migration.
The management server and the service server of the said node location that mediates are provided with nodes encoding;
Said nodes encoding has uniqueness;
Said nodes encoding is elongated, is made up of 0,1 string, and its nodes encoding length equals the common prefix length partly of the server node coding or the service node coding of its management; The nodes encoding of management server is the common prefix part of nodes encoding of nodes encoding or the service server of its subordinate's management server; The nodes encoding of service server is the common prefix part of the service node coding of its management;
The length of service server nodes encoding is zero when only containing a service server.
Said management server is equipped with the prefix bit tabulation and preserves the management server of its management or the nodes encoding of service server, corresponding management server of each prefix bit string or service server;
Said service server is preserved a prefix bit string, manages all service node codings with this prefix bit string beginning;
Said system also comprises service node, and said service node is represented user load, is provided with the service node coding, and said service node coding is a fixed length, length overall T position, and service node coding in said T position comprises the K position nodes encoding prefix of upper level service server,
Said K, T are nonnegative integer.
Said management server and service server are equipped with and merge threshold value and division threshold value, and the division threshold value is greater than merging threshold value.
Also be provided with refusal sign r on the said management server, the value of r is a true or false, and said r is that true time is represented to reach stable state by the management server or the service server of this management service management, can not divide.
To put genuine condition following for refusal sign r on the said management server:
(1) is positioned at i layer and 0<i<the refusal sign r of the management server of n puts genuine condition and is: the r that is positioned at father's node is for true, and the prefix bit string length in the prefix bit tabulation is
L
sBe the maximum node code length on the s layer-management server logic, said this non-service server of management server lower floor;
(2) r that is arranged in the management server of root position puts genuine condition and is: the prefix bit string length of prefix bit tabulation is L
1, said this non-service server of management server lower floor, said L
1The maximum node code length of expression ground floor management server;
(3) r that is positioned at n layer-management server puts genuine condition and is: the r of father's node is for true, and the prefix bit string length in the prefix bit tabulation is M, and said M is the maximum node code length on the traffic server logic, and promptly lower floor's service server is full,
Said i, n, M are nonnegative integer.
The member management module of said management server or service server adopts following steps to accomplish division and merging;
Said step toward division is: the monitoring thread of management server or service server is sent the division request message to higher level's management server; The monitoring thread of upper management server is handled the division request based on the division condition of management server or service server, and is issued the division response instruction behind the division request message that receives the transmission of subordinate's server; After the monitoring thread of subordinate's server was received the division response instruction, the division thread divided based on the division instruction, and local server is carried out load reallocation and business migration;
Said combining step is: the monitoring thread of management server is according to the management server of the subordinate's server that monitors or the running status of service server, and sends merge command according to the merging condition of management server or service server to the server that reaches the merging condition; After the monitoring thread of subordinate's management server or service server is received and merged response instruction, merge thread and merge, and local server is carried out load merging and business migration according to merge command;
Service Processing Module in the said service server is managed business according to service request from user:
Be specially: the user sends Business Processing message to the Service Processing Module of known business server; The Service Processing Module of service server judges according to the service node coding whether the user belongs to this service server after the Business Processing message that receives user's transmission, if; Then carry out Business Processing; Otherwise, then sending operation assistance and handle request until finding the corresponding business service device to higher level's management server, the corresponding business service device is received the laggard industry affair of Business Processing message processing.
Said division condition comprises service server division condition and management server division condition;
Said service server division condition: the service node coding of management surpasses the division threshold value;
Said management server division condition is to satisfy one of following condition:
(1) division at the same level, the refusal sign r of this management server is false, top management server r is that the server of vacation and top management server management reaches the upper limit, then division, the management server after the division is managed by top management server;
(2) merge division; The refusal sign r of this management server is false; Top management server r is that the server of vacation and top management server management reaches the upper limit, then division, and newly-increased management service management has common prefix bit string and server node code length greater than L
cManagement server, the nodes encoding length of newly-increased management server equals L
c
(3) upwards division, the refusal sign r of this management server is false, top management server r is true, division immediately, and the server after the newly-increased management service management division;
The said management server that is positioned at the root position adopts upwards division.
Said merging condition comprises that service server merges condition and management server merges condition;
Said service server merging condition is: the load sum of the two-server that the service server nodes encoding is adjacent is lower than the merging threshold value;
Management server merging condition is to satisfy one of following condition:
(1) level merges, said two management server s with layer
IjWith s
IkAdjacent, the load sum of nodes encoding be lower than merge threshold value and the child nodes type identical;
(2) first vertically merge management server s
Ij, s
IjChild node be server s
(i+1) kAnd s
(i+1) kChild nodes be management server, s
IjLoad and s
(i+1) kThe load sum be lower than and merge threshold value or s
IjLoad and s
(i+1) kThe load sum make s
IjThe refusal flag bit be true, with s
(i+1) kMerge to s
Ij
(3) second vertically merge management server s
Ij, management server s
(i+1) kBe s
IjChild node and unique, s
(i+1) kHas only unique child s
(i+2) kAnd s
(i+2) kBe service server, with s
(i+1) kMerge to s
Ij
Wherein, s
IjRepresent i layer, j platform management server; In the merging process of management server, horizontal direction merges will have precedence over the vertical direction merging; The vertical direction that merges is bottom-up; Server after level merges is replaced by the many servers of former load.
Beneficial effect of the present invention:
The present invention is divided into member management module and operation assistance processing module with management server, and service server is divided into member management module and Service Processing Module; By the operation assistance processing module of management server and the Service Processing Module of service server service node is carried out Business Processing; The member management module of management server is according to the service server that listens to/or the running status of next stage management server and judge whether needs merge; Merge if desired, the member management module of management server is organized service server/or the merging and the load reallocation of next stage management server; The member management module of the service server/management server one-level management server that when load is overweight, makes progress sends division request; The load reallocation and the balance of next stage server accomplished in the member management intermodule cooperation of management server; Realized the monitoring of service operation state; Can check the service operation state easily, and then manage efficiently, and make the server zone architecture system have the balance of higher scalability and framework; Load migration moves between adjacent server as far as possible simultaneously, can reduce the resource consumption that load migration brings.
Description of drawings
Fig. 1 is a Scalable Network Services device gang fight construction system structure chart;
Fig. 2 is service node coding structure figure of the present invention;
The division preceding system construction drawing of Fig. 3 (a) for only relating to service server in the embodiment of the invention (b) is the system construction drawing after the division that only relates to service server in the embodiment of the invention;
Fig. 4 (a) (b) is the system construction drawing after the service server division that causes the division of top service device in the embodiment of the invention for causing the preceding system construction drawing of service server division of top service device division in the embodiment of the invention;
Fig. 5 (a) is the system construction drawing before the peer's division in the embodiment of the invention, (b) is the system construction drawing after the peer's division in the embodiment of the invention;
Fig. 6 (a) is the system construction drawing before the upwards division in the embodiment of the invention, (b) is the upwards division back system construction drawing in the embodiment of the invention;
Fig. 7 (a) is a system construction drawing before the merging division in the embodiment of the invention, (b) is the merging division back system construction drawing in the embodiment of the invention;
Fig. 8 (a) is a system construction drawing before the level in the embodiment of the invention merges, and (b) is that the level in the embodiment of the invention merges the back system construction drawing;
Fig. 9 (a) is the preceding system construction drawing of the first vertical merging in the embodiment of the invention, (b) is the first vertical back system construction drawing that merges in the embodiment of the invention;
Figure 10 (a) is the preceding system construction drawing of the second vertical merging in the embodiment of the invention, (b) is the system construction drawing after (a) figure level merges, and (c) is the system construction drawing of (b) figure after second vertical the merging.
Embodiment
Below in conjunction with embodiment and accompanying drawing, the present invention is done to specify further, but execution mode of the present invention is not limited thereto.
Embodiment
As shown in Figure 1, the Scalable Network Services device gang fight structure in the present embodiment comprises several management servers and service server, and said Scalable Network Services device gang fight structure is tree-like framework;
Said management server connects several management servers or several service servers; Said management server is positioned at the root node or the intermediate node position of tree-like framework, and service server is positioned at the leaf node position of tree-like framework;
Said management server comprises member management module and operation assistance processing module;
Be positioned at the member management module on the management server of root node, be used for when load surpasses the division threshold value, sending the division request to the keeper; Be used to monitor the running status of subordinate's server,, then issue merge command to these servers if subordinate's server reaches the merging condition; Be used for after the division request that receives subordinate's server transmission, handling the division of subordinate's server; Be used for the division and the load reallocation of this machine of processing server; Be used to handle merging with subordinate's management server;
Be positioned at the member management module on the management server of intermediate node, be used to receive the division response that the upper management server issues after, carry out load reallocation and business migration; Be used for after receiving the merge command that the upper management server issues, carrying out load merging and business migration; Be used to monitor the running status of local server, state information be saved in database, and with the state information timed sending to higher level's management server; Be used for when load surpasses the division threshold value, sending the division request to higher level's server; Be used to monitor the running status of subordinate's server,, then issue merge command to these servers if subordinate's server reaches the merging condition; Be used for after the division request that receives subordinate's server transmission, handling the division of subordinate's server; Be used for the nodes encoding management; Be used to refuse flag management;
Said operation assistance processing module, the service server under the user is searched in the operation assistance processing request that the operation assistance processing is asked or subordinate's management server is transmitted that is used for sending according to service server
Comprise Service Processing Module and member management module on the said service server;
Member management module on the service server, be used to receive the division response that the upper management server issues after, carry out load reallocation and business migration; Be used for after receiving the merge command that the upper management server issues, carrying out load merging and business migration; Be used to monitor the running status of local server, state information be saved in database, and with the state information timed sending to higher level's management server; Be used for when load surpasses the division threshold value, sending the division request to higher level's server; Be used for the nodes encoding management;
Said Service Processing Module is used for interface to the user being provided, and manages business according to service request from user.
Member management module in the said management server comprises monitoring thread, division thread and merges thread;
Member management module in the service server comprises monitoring thread, division thread and merges thread;
Monitoring thread on the said management server that is positioned at root node is used to monitor the running status of local server and state information is saved in database; Be used to monitor the running status of subordinate's server, be used for sending the division request, send merge command to the subordinate's server that reaches the merging condition to the keeper; Be used for nodes encoding management and refusal flag management;
Monitoring thread on the said management server that is positioned at intermediate node is used to monitor the running status of local server, and state information is saved in database, and with the supreme one-level management server of state information timed sending; Be used to monitor the running status of subordinate's server, be used for sending the division request, send merge command to the subordinate's server that reaches the merging condition to higher level's server; Receive division response instruction and merging response instruction that the upper level server sends; Be used for nodes encoding management and refusal flag management;
Monitoring thread on the said service server is used to monitor the running status of local server, and state information is saved in database, and with the state information timed sending to higher level's management server; Be used for when load surpasses the division threshold value, upwards the one-level management server sends the division request; Be used to receive the division response instruction that the upper management server sends; Be used to receive the merge command that the upper management server sends; Be used for the nodes encoding management;
Division thread in said management server and the service server is used to handle load reallocation and business migration;
Merging thread in said management server and the service server is used to handle load merging and business migration.
Adopt TCP to communicate between the member management module on the member management module of said management server and the member management module of management server or the service server; Adopt TCP to communicate between operation assistance processing module on the management server and operation assistance processing module on the management server or the Service Processing Module on the service server; Between the operation assistance processing module and member management module on the management server, communicate through the native call mode; Between Service Processing Module on the service server and the member management module, communicate through the native call mode.
The management server of the said node location that mediates is provided with nodes encoding to distinguish different servers with service server, and nodes encoding has uniqueness;
Said nodes encoding is elongated, is made up of 0,1 string, and said nodes encoding length equals the server node coding of its management or the common prefix partial-length of service node coding; The length of service server coding is zero when server zone only contains a service server; The code length of the management server of root position is zero.
Management server is equipped with the prefix bit tabulation and preserves the management server of its management or the nodes encoding of service server, corresponding management server of each prefix bit string or service server;
Said service server is preserved a prefix bit string, manages all service node codings with this prefix bit string beginning
In case study on implementation of the present invention, the server zone initial condition includes only a management server and a service server, and the server node code length is zero (need not coding); The nodes encoding prefix bit of management server is tabulated to empty, and r is false, and the nodes encoding prefix bit string length of service server is zero, and service server directly receives the also service request of process user.
As shown in Figure 1, tree-like framework is three layers in the present embodiment, and the 1st layer has two management servers, and the nodes encoding prefix bit tabulation that then is positioned at the management server of root position comprises (0,1), and the server node coding of management is respectively 0 and 1; Server node is encoded under 0 the management server has three service servers, and the tabulation of nodes encoding prefix bit is (00,010,011), and the service server nodes encoding of management is remembered respectively and done 00,010 and 011; Server node is encoded under 1 the management server has two service servers, and the tabulation of nodes encoding prefix bit is (10,11), and the service server nodes encoding of management is remembered respectively and done 10 and 11.
As shown in Figure 1, the Scalable Network Services device gang fight structure of present embodiment is a tree height of three, and 0 among the figure, 1 string list show the nodes encoding of each server.
The running of the Scalable Network Services device gang fight structure shown in Fig. 1 is:
(1) disposes management server and startup;
(2) member management module and the operation assistance processing module on the deployment management server.
(3) dispose service server and startup;
(4) member management module and the Service Processing Module on the deployment service server;
(5) relationship between superior and subordinate of management server and service server is set;
Aforesaid operations can have been managed Scalable Network Services device crowd system after accomplishing.
Be illustrated in figure 2 as service node coding structure figure of the present invention, said service node encoding setting is on service node, and service node is represented user load, supposes that management server can manage 2 at most in logic
LcIndividual management server or 2
MIndividual service server, service server can support 2 at most
NIndividual service node coding, the server zone framework has the n layer to be positioned at the management server of intermediate node at most; Access server crowd's service node has corresponding with it unique service node coding, and this service node coding is confirmed the service server that service node inserts in server zone; The service node coding is a fixed length, and length overall T position is made up of the local service node coding of K position server node coding prefix and N position, and K position server node coding prefix is by L position management server coding (L≤n*L
c, L
cBe nonnegative integer) and M position service server coded portion composition; Wherein, M, N, n, T, K, L are nonnegative integer.
Said every station server is equipped with and merges threshold value and division threshold value, and the division threshold value is the maximum load that server can bear, and merges threshold value and is meant the maximum load that merges the back server, and the division threshold value is usually greater than merging threshold value.Wherein, also be provided with refusal sign r on the management server, the value of r is a true or false, and r is that true time representes that the management server/service server of server admin reaches stable state, can not divide.
It is following that said refusal sign r puts genuine condition:
(1) is positioned at i layer and 0<i<n, the r of management server put genuine condition and be: the r of father's node is for true, and the prefix bit string length in the prefix bit tabulation is
L
sBe the maximum code length on the s layer-management server logic, the non-service server of lower floor;
(2) r that is arranged in the management server of root position puts genuine condition and is: the prefix bit string length of prefix bit tabulation is L
1, the non-service server of lower floor, said L
1The maximum node code length of expression ground floor management server;
(3) r that is positioned at n layer-management server puts genuine condition and is: the r of father's node is for true, and the prefix bit string length in the prefix bit tabulation is M, and said M is the maximum node code length on the traffic server logic, and promptly lower floor's service server is full.Said i, n, M are nonnegative integer.
The member management module of said management server or service server adopts following steps to accomplish division and merging:
Said step toward division is: the monitoring thread of management server or service server is sent the division request message to higher level's management server; The monitoring thread of upper management server is handled the division request message according to the division condition of management server or service server, and is issued the division response instruction behind the division request message that receives the transmission of subordinate's server; After the monitoring thread of subordinate's server was received the division response instruction, the division thread divided according to the division instruction, and local server is carried out load reallocation and business migration;
Said combining step is: the monitoring thread of management server is according to the subordinate's management server that monitors or the running status of service server, and sends merge command according to the merging condition of management server or service server to the server that reaches the merging condition; After the monitoring thread of subordinate's server is received merge command, merge thread and merge, and local server is carried out load merging and business migration according to merge command;
Service Processing Module in the said service server is managed business according to service request from user; Be specially: the user sends Business Processing message to the Service Processing Module of known business server; The Service Processing Module of service server judges according to the service node coding whether the user belongs to this service server after the Business Processing message that receives user's transmission, if; Then carry out Business Processing; Otherwise, then sending operation assistance and handle request until finding the corresponding business service device to higher level's management server, the corresponding business service device is received the laggard industry affair of Business Processing message processing.
The condition of said service server division is: the service node of management surpasses the division threshold value.
For explaining conveniently, suppose that every layer service server can manage four nodes, i.e. Lc=2 at most; In other embodiments of the invention, the server zone framework before the division can be a multilayer.
Fig. 3 (a) and (b) are the forward and backward system construction drawings of division that only relate to service server in the embodiment of the invention.Shown in Fig. 3 (a), before the division, service server has three; The nodes encoding prefix bit tabulation that is positioned at the management server of root position is (0,10,11); The refusal sign r of the management server of root position is false, and the server node coding of management is respectively 0,10 and 11; When the load of service server surpassed the division threshold value, the monitoring thread of service server was sent the division request to higher level's management server; The division thread of upper management server receives and handles the division of subordinate's server, and shown in Fig. 3 (b), nodes encoding prefix bit string is that 11 service server splits into two; The newly-increased server of the use that load is few; Two service server nodes encodings are remembered work 110 and 111 respectively, if after the division, the load of service server still surpasses the division threshold value; Continue this moment upwards to send the division request, be no more than the division threshold value until server load; The nodes encoding prefix bit tabulation note of upper management server is done (0,10,110,111), and r is false, the nodes encoding of server constant (management server of root position does not have nodes encoding in this example).
Fig. 4 (a) and (b) are the service server division front and back system construction drawings that cause the top management server division in the embodiment of the invention.Shown in Fig. 4 (a), before the division, tree-like framework is two-layer, and service server has four; The nodes encoding prefix bit tabulation of supposing to be positioned at the management server of root position comprises (0,10,110; 111), refusal sign r is false, and the service server nodes encoding is respectively 0,10,110 and 111; When the load of service server surpasses the division threshold value (like the prefix bit string be 111 service server), the monitoring thread of service server is sent the division request to higher level's management server; The upper management server receives and handles the division of subordinate's server.
Be depicted as the Organization Chart after the division like Fig. 4 (b), tree-like framework is three layers, and said nodes encoding prefix bit is that 111 service server splits into two, the newly-increased server of the use that load is few, and the distribution server nodes encoding is remembered respectively and is done 1110 and 1111; Because service server expands to five by four, the monitoring thread that then is positioned at the management server of root position is sent the division request to the manager; The manager receives the division request; The management server that indication is positioned at the root position is split into 0 and 1 (one newly-increased, and a use is positioned at the former server of root position, the newly-increased server of the use that load is few); A newly-increased station server is positioned at root node, the server after the management division; The nodes encoding prefix bit tabulation that is positioned at the management server of root position comprises (0,1), and r is false; Service server moves according to the nodes encoding prefix bit string of self, the prefix bit string be 0 service server to migrate to nodes encoding be that the nodes encoding prefix bit tabulation of management server comprises (0) under 0 the management server, r be vacation; The nodes encoding prefix bit tabulation that is encoded to 1 management server comprises (10,110,1110,1111), and r be vacation.
If after the division, the load of server still surpasses the division threshold value, continues this moment upwards to send the division request, is no more than the division threshold value until server load.
Must satisfy one of following condition during the management server division:
(1) division at the same level, r is false, top service device r does not reach the upper limit for physical server false and management, division immediately, the server after the division is by the top service management.
Shown in Fig. 5 (a) before the division tree-like framework be three layers, the 1st layer has two management servers, the nodes encoding prefix bit tabulation that then is positioned at the management server of root position is (0,1), r be false, the server node coding of management is note work 0 and 1 respectively; Server node is encoded under 0 the management server has two service servers, and the nodes encoding prefix bit tabulation of management server is (00,01), and r be vacation; Server node is encoded under 1 the management server has four service servers, and the nodes encoding prefix bit tabulation of management server is (100,101,110,111), and r be vacation; When the load of service server surpassed the division threshold value, the monitoring thread of service server was sent the division request to higher level's management server; The upper management server receives and handles the division of subordinate's server.
Be depicted as the system construction drawing after the division like Fig. 5 (b); Said tree-like framework is three layers; Like the nodes encoding prefix bit is that 100 service server splits into two, the newly-increased server of the use that load is few, and upper management server-assignment server node coding note respectively does 1000 and 1001; Because service server expands to five by four, then server node is encoded to 1 management server and sends the division request to higher level's management server; The management server that is positioned at the root position receives and handles the division of subordinate's server; Server node is encoded to 1 management server and splits into two; The newly-increased server of the use that load is few, the management server distribution server nodes encoding that is positioned at the root position is remembered work 10 and 11 respectively; The nodes encoding prefix bit tabulation that is positioned at the management server of root position is (0,10,11), and r is false; Service server moves according to the nodes encoding prefix bit string of self; The prefix bit string is that 110,111 service server migrates to server node and is encoded under 11 the management server; The nodes encoding prefix bit tabulation of management server is (110,111), and r is false; The nodes encoding prefix bit tabulation that is encoded to 10 management server comprises (1000,1001,101); R is false (if after the division; The load of server still surpasses the division threshold value, continues upwards to send the division request this moment, is no more than the division threshold value until server load).
(2) merge division, the refusal sign r of said management server be false, and top service device r is for physical server false and management reaches the upper limit, divides immediately and newly-increased management service management has common prefix bit string and server node code length greater than L
cManagement server, the nodes encoding length of newly-increased management server equals L
c
Fig. 7 (a) and (b) are to merge the forward and backward system construction drawing of division in the embodiment of the invention.
Be depicted as the system construction drawing before the division like Fig. 7 (a), shown in tree-like framework be three layers, when said load when service server surpassed the division threshold value, the monitoring thread of service server was sent division to higher level's management server and is asked; The upper management server receives and handles the division of subordinate's server.
Shown in Fig. 7 (b), be that 11111 service server splits into two like nodes encoding prefix bit string, the newly-increased server of the use that load is few, upper management server-assignment server node coding note respectively does 111110 and 111111; Because service server expands to five by four, nodes encoding is that the monitoring thread of 111 management server is sent the division request to higher level's management server (be root in this example, other situation are similar); The upper management server receives and handles the division of subordinate's management server, and nodes encoding is that 111 management server splits into two, the newly-increased server of the use that load is few, and upper management server-assignment server node coding note respectively does 1110 and 1111; Because the 1st layer-management server is that physical server false and management reaches the upper limit by four r that expand to five and root, needing a newly-increased nodes encoding is 11 management server, and the tabulation of nodes encoding prefix bit comprises (110,1110,1111), and r be vacation.
If being 0 management server, nodes encoding splits into two, promptly 00,01, and then newly-increased management server nodes encoding is 11, the tabulation of nodes encoding prefix bit comprises (110,111); If being 10 management server, nodes encoding splits into two; Promptly 100,101, then newly-increased two management servers, nodes encoding are remembered respectively and are done 10,11; The prefix bit tabulation that wherein is encoded to 10 management server comprises (100; 101) the prefix bit tabulation that, is encoded to 11 management server comprises (110,111).
The nodes encoding of three station servers of management is respectively 110,1110,1111; Service server moves according to the nodes encoding prefix bit string of self; The prefix bit string is that to migrate to nodes encoding be under 1110 the management server for 11100,11101 service server; The nodes encoding prefix bit tabulation of management server comprises (11100,11101), and r is false; The nodes encoding prefix bit tabulation that is encoded to 1111 management server comprises (11110,111110,111111); R is false, if after the division, the load of server still surpasses the division threshold value; Continue this moment upwards to send the division request, be no more than the division threshold value until server load.
(3) upwards division, the refusal sign r of said management server is false, top management server r is true, division immediately, and the server after the newly-increased management service management division; The management server that is positioned at the root position adopts upwards division.Wherein, in the horizontal split process, the newly-increased server of the use that load is few usually.
Fig. 6 (a) (b) is depicted as in the present embodiment the upwards forward and backward system construction drawing of division.
Shown in Fig. 6 (a) before the division the tree-like framework of server zone be three layers, server node is encoded under 11 the management server has four service servers; When the load of service server surpassed the division threshold value, the monitoring thread of service server was sent the division request to higher level's management server; The upper management server receives and handles the division of subordinate's server.
System construction drawing shown in Fig. 6 (b) after the division is that 1111 service server splits into two like the nodes encoding prefix bit, the newly-increased server of the use that load is few, and upper management server-assignment server node coding note respectively does 11110 and 11111; Because service server expands to five by four, the monitoring thread that server node is encoded to 11 management server is sent division to higher level's management server (be the management server of root position in this example, other situation are similar) asks.
The refusal sign r of root position is true; Can't admit new server; Instructs node is encoded to 11 management server and is split into 110 and 111 (one newly-increased; A use is encoded to former 11 server, the newly-increased server of the few use of load), newly-increased one is encoded to 110 and 111 after 11 the server admin division; The nodes encoding prefix bit tabulation that is encoded to 11 management server is (110,111), and r be vacation.
Service server moves according to the nodes encoding prefix bit string of self; The prefix bit string is that to migrate to nodes encoding be under 110 the management server for 1100,1101 service server; The nodes encoding prefix bit tabulation of management server comprises (1100,1101), and r is false; The nodes encoding prefix bit tabulation that is encoded to 111 management server comprises (1110,11110,11111); R is false, if after the division, the load of server still surpasses the division threshold value; Continue this moment upwards to send the division request, be no more than the division threshold value until server load.
Can know that by above-mentioned example the cleavage direction of server is downward the end of from; If after the division, the load of server still surpasses the division threshold value, then continues to send the division request to higher level's server, is no more than the division threshold value until the load of server.
Said merging condition comprises that service server merges condition and management server merges condition;
Said service server merging condition is: the load sum of the two-server that the service server nodes encoding is adjacent is lower than the merging threshold value;
Management server merging condition is to satisfy one of following condition:
(1) level merges, said two management server s with layer
IjWith s
IkAdjacent, the load sum of nodes encoding be lower than merge threshold value and the child nodes type identical;
(2) first vertically merge management server s
Ij, s
IjChild node be server s
(i+1) kAnd s
(i+1) kChild nodes be management server, s
IjLoad and s
(i+1) kThe load sum be lower than and merge threshold value (perhaps s
IjLoad and s
(i+1) kThe load sum make s
IjThe refusal flag bit be true), with s
(i+1) kMerge to s
Ij
(3) second vertically merge management server s
Ij, management server s
(i+1) kBe s
IjChild node and unique, s
(i+1) kHas only unique child s
(i+2) kAnd s
(i+2) kBe service server, with s
(i+1) kMerge to s
Ij
Wherein, s
IjRepresent i layer, j platform management server; In the merging process of management server, horizontal direction merges will have precedence over the vertical direction merging; The vertical direction that merges is bottom-up; Server after level merges is replaced by the many servers of former load.
Fig. 8 is that the level in the present embodiment merges forward and backward Organization Chart.Wherein, the server zone framework before Fig. 8 (a) expression merges, the server zone framework after Fig. 8 (b) expression merges.The top service device (is root node in this example; Other situation are similar) the monitoring thread running status that listens to two-server with adjacent node coding (be 10 and 11 management server like nodes encoding in this example; Lower floor also can be a service server in other instances); The load sum of two-server is no more than the merging threshold value, judges that two-server can merge; The upper management server sends merge command to the two-server that subordinate need merge; Subordinate's server receives and handles merge command; Two-server is merged into a station server; (many in this example on the server of business migration on the d/d server after merge by load; Be that origin node is encoded to 11 management server replacement, the management server nodes encoding after the merging is 1; The prefix bit string be 10 service server to migrate to nodes encoding be that the nodes encoding prefix bit tabulation of the management server after the merging comprises (10,110,111) under 1 the management server, r be vacation).
Fig. 9 (a) is the forward and backward Organization Chart of the vertical merging of first in the present embodiment (b).Wherein, the server zone framework before Fig. 9 (a) expression merges, the server zone framework after Fig. 9 (b) expression merges.As above stratum server (is root node in this example; Other situation are similar) monitoring thread listen to the running status that nodes encoding is 1 management server; It is management server that the load of top service device and the load sum that is encoded to 1 management server are no more than the child node that merges threshold value and be encoded to 1 management server; Satisfy the vertical condition that merges (1), can vertically merge; Top management server is sent merge command to the server that subordinate need merge; The subordinate server receives also handles merge command, is encoded to 1 management server and merges to top service device (be root node in this example, other situation are similar); (the prefix bit string is that 10,11 management server migrates under the management server that is positioned at the root position on the server of business migration on the d/d server after merge; The nodes encoding prefix bit tabulation of the management server after the merging comprises (0; 10,11), r is false.
Figure 10 (a) (b) (c) is the forward and backward Organization Chart of the vertical merging of second in the present embodiment.Wherein, the server zone framework before Figure 10 (a) expression merges, the server zone framework after Figure 10 (c) expression merges.As above stratum server (is root node in this example; Other situation are similar) monitoring thread be encoded to the running status of 1 management server according to listening to running status that nodes encoding is 0 management server and adjacent node; The load sum of two-server is no more than the merging threshold value; Judge that two-server can merge (level merging), shown in Figure 10 (b).The upper management server sends merge command to two management servers that subordinate need merge; The subordinate management server receives also handles merge command, and two-server is merged into one and many by load, and promptly origin node is encoded to 0 management server and replaces.
Because after merging; The upper management server (is root node in this example; Other situation are similar) under management server have only the child nodes of and lower floor's management server to be service server, then satisfy the second vertical condition that merges, carry out second and vertically merge; The levels server is merged into a management server and is replaced server node coding constant (root does not have nodes encoding in this example) by lower floor's management server; The nodes encoding prefix bit tabulation of management server comprises (00,01,1), and r is false, shown in Figure 10 (c).
By above-mentioned visible, the present invention is divided into member management module and operation assistance processing module with management server, and service server is divided into member management module and Service Processing Module; By the operation assistance processing module of management server and the Service Processing Module of service server service node is carried out Business Processing; The member management module of management server is according to the service server that listens to/or the running status of next stage management server and judge whether needs merge; Merge if desired, the member management module of management server is organized service server/or the merging and the load reallocation of next stage management server; The member management module of the service server/management server one-level management server that when load is overweight, makes progress sends division request; The load reallocation and the balance of next stage server accomplished in the member management intermodule cooperation of management server; Realized the monitoring of service operation state; Can check the service operation state easily, and then manage efficiently, and make the server zone architecture system have the balance of higher scalability and framework; Load migration moves between adjacent server as far as possible simultaneously, can reduce the resource consumption that load migration brings.
The foregoing description is a preferred implementation of the present invention; But execution mode of the present invention is not limited by the examples; Other any do not deviate from change, the modification done under spirit of the present invention and the principle, substitutes, combination, simplify; All should be the substitute mode of equivalence, be included within protection scope of the present invention.
Claims (10)
1. a Scalable Network Services device gang fight construction system is characterized in that, comprises several management servers and service server, and said Scalable Network Services device gang fight structure is tree-like framework;
Said management server connects several management servers or several service servers; Said management server is positioned at the root node or the intermediate node position of tree-like framework, and service server is positioned at the leaf node position of tree-like framework;
Said management server comprises member management module and operation assistance processing module;
Member management module on the said management server that is positioned at root node is used for when load surpasses the division threshold value, sending the division request to the keeper; Be used to monitor the running status of subordinate's server,, then issue merge command to these servers if subordinate's server reaches the merging condition; Be used for after the division request that receives subordinate's server transmission, handling the division of subordinate's server; Be used for the division and the load reallocation of this machine of processing server; Be used to handle merging with subordinate's management server;
Member management module on the said management server that is positioned at intermediate node, be used to receive the division response that the upper management server issues after, carry out load reallocation and business migration; Be used for after receiving the merge command that the upper management server issues, carrying out load merging and business migration; Be used to monitor the running status of local server, state information be saved in database, and with the state information timed sending to higher level's management server; Be used for when load surpasses the division threshold value, sending the division request to higher level's management server; Be used to monitor the running status of subordinate's server,, then issue merge command to these servers if subordinate's server reaches the merging condition; Be used for after the division request that receives subordinate's server transmission, handling the division of subordinate's server; Be used for the nodes encoding management; Be used to refuse flag management;
Said operation assistance processing module, the service server under the user is searched in the operation assistance processing request that the operation assistance processing is asked or subordinate's management server is transmitted that is used for sending according to service server;
Comprise Service Processing Module and member management module on the said service server;
Member management module on the said service server, be used to receive the division response that the upper management server issues after, carry out load reallocation and business migration; Be used for after receiving the merge command that the upper management server issues, carrying out load merging and business migration; Be used to monitor the running status of local server, state information be saved in database, and with the state information timed sending to higher level's management server; Be used for when load surpasses the division threshold value, sending the division request to higher level's server; Be used for the nodes encoding management;
Said Service Processing Module is used for interface to the user being provided, and manages business according to service request from user.
2. Scalable Network Services device gang fight construction system according to claim 1 is characterized in that, the member management module in the said management server comprises monitoring thread, division thread and merges thread;
Member management module in the said service server comprises monitoring thread, division thread and merges thread;
Monitoring thread on the said management server that is positioned at root node is used to monitor the running status of local server and state information is saved in database; Be used to monitor the running status of subordinate's server, be used for sending the division request, send merge command to the subordinate's server that reaches the merging condition to the keeper; Be used for nodes encoding management and refusal flag management;
Monitoring thread on the said management server that is positioned at intermediate node is used to monitor the running status of local server, and state information is saved in database, and with the supreme one-level management server of state information timed sending; Be used to monitor the running status of subordinate's server, be used for sending the division request, send merge command to the subordinate's server that reaches the merging condition to higher level's server; Receive division response instruction and merging response instruction that the upper level server sends; Be used for nodes encoding management and refusal flag management;
Monitoring thread on the said service server is used to monitor the running status of local server, and state information is saved in database, and with the state information timed sending to higher level's management server; Be used for when load surpasses the division threshold value, upwards the one-level management server sends the division request; Be used to receive the division response instruction that the upper management server sends; Be used to receive the merge command that the upper management server sends; Be used for the nodes encoding management;
Division thread in said management server and the service server is used to handle load reallocation and business migration;
Merging thread in said management server and the service server is used to handle load merging and business migration.
3. Scalable Network Services device gang fight construction system according to claim 1 is characterized in that the management server and the service server of the said node location that mediates are provided with nodes encoding;
Said nodes encoding has uniqueness;
Said nodes encoding is elongated, is made up of 0,1 string, and its nodes encoding length equals the common prefix length partly of the server node coding or the service node coding of its management; The nodes encoding of management server is the common prefix part of nodes encoding of nodes encoding or the service server of its subordinate's management server; The nodes encoding of service server is the common prefix part of the service node coding of its management;
The length of service server nodes encoding is zero when only containing a service server.
4. Scalable Network Services device gang fight construction system according to claim 1; It is characterized in that; Said management server is equipped with the prefix bit tabulation and preserves the management server of its management or the nodes encoding of service server, corresponding management server of each prefix bit string or service server;
Said service server is preserved a prefix bit string, manages all service node codings with this prefix bit string beginning;
Said system also comprises service node, and said service node is represented user load, is provided with the service node coding, and said service node coding is a fixed length, length overall T position, and service node coding in said T position comprises the K position nodes encoding prefix of upper level service server,
Said K, T are nonnegative integer.
5. Scalable Network Services device gang fight construction system according to claim 1 is characterized in that, said management server and service server are equipped with and merge threshold value and division threshold value, and the division threshold value is greater than merging threshold value.
6. Scalable Network Services device gang fight construction system according to claim 4; It is characterized in that; Also be provided with refusal sign r on the said management server; The value of r is a true or false, and said r is that true time is represented to reach stable state by the management server or the service server of this management service management, can not divide.
7. Scalable Network Services device gang fight construction system according to claim 6 is characterized in that, to put genuine condition following for refusal sign r on the said management server:
(1) is positioned at i layer and 0<i<the refusal sign r of the management server of n puts genuine condition and is: the r that is positioned at father's node is for true, and the prefix bit string length in the prefix bit tabulation is
L
sBe the maximum node code length on the s layer-management server logic, said this non-service server of management server lower floor;
(2) r that is arranged in the management server of root position puts genuine condition and is: the prefix bit string length of prefix bit tabulation is L
1, said this non-service server of management server lower floor, said L
1The maximum node code length of expression ground floor management server;
(3) r that is positioned at n layer-management server puts genuine condition and is: the r of father's node is for true, and the prefix bit string length in the prefix bit tabulation is M, and said M is the maximum node code length on the traffic server logic, and promptly lower floor's service server is full,
Said i, n, M are nonnegative integer.
8. like each described Scalable Network Services device gang fight construction system of claim 1-7, it is characterized in that,
The member management module of said management server or service server adopts following steps to accomplish division and merging;
Said step toward division is: the monitoring thread of management server or service server is sent the division request message to higher level's management server; The monitoring thread of upper management server is handled the division request based on the division condition of management server or service server, and is issued the division response instruction behind the division request message that receives the transmission of subordinate's server; After the monitoring thread of subordinate's server was received the division response instruction, the division thread divided based on the division instruction, and local server is carried out load reallocation and business migration;
Said combining step is: the monitoring thread of management server is according to the management server of the subordinate's server that monitors or the running status of service server, and sends merge command according to the merging condition of management server or service server to the server that reaches the merging condition; After the monitoring thread of subordinate's management server or service server is received and merged response instruction, merge thread and merge, and local server is carried out load merging and business migration according to merge command;
Service Processing Module in the said service server is managed business according to service request from user:
Be specially: the user sends Business Processing message to the Service Processing Module of known business server; The Service Processing Module of service server judges according to the service node coding whether the user belongs to this service server after the Business Processing message that receives user's transmission, if; Then carry out Business Processing; Otherwise, then sending operation assistance and handle request until finding the corresponding business service device to higher level's management server, the corresponding business service device is received the laggard industry affair of Business Processing message processing.
9. Scalable Network Services device gang fight construction system according to claim 8 is characterized in that, said division condition comprises service server division condition and management server division condition;
Said service server division condition: the service node coding of management surpasses the division threshold value;
Said management server division condition is to satisfy one of following condition:
(1) division at the same level, the refusal sign r of this management server is false, top management server r is that the server of vacation and top management server management reaches the upper limit, then division, the management server after the division is managed by top management server;
(2) merge division; The refusal sign r of this management server is false; Top management server r is that the server of vacation and top management server management reaches the upper limit; Then division, newly-increased management service management have common prefix bit string and the server node code length management server greater than Lc, and the nodes encoding length of newly-increased management server equals Lc;
(3) upwards division, the refusal sign r of this management server is false, top management server r is true, division immediately, and the server after the newly-increased management service management division;
The said management server that is positioned at the root position adopts upwards division.
10. Scalable Network Services device gang fight construction system according to claim 8 is characterized in that, said merging condition comprises that service server merges condition and management server merges condition;
Said service server merging condition is: the load sum of the two-server that the service server nodes encoding is adjacent is lower than the merging threshold value;
Management server merging condition is to satisfy one of following condition:
(1) level merges, said two management server s with layer
IjWith s
IkAdjacent, the load sum of nodes encoding be lower than merge threshold value and the child nodes type identical;
(2) first vertically merge management server s
Ij, s
IjChild node be server s
(i+1) kAnd s
(i+1) kChild nodes be management server, s
IjLoad and s
(i+1) kThe load sum be lower than and merge threshold value or s
IjLoad and s
(i+1) kThe load sum make s
IjThe refusal flag bit be true, with s
(i+1) kMerge to s
Ij
(3) second vertically merge management server s
Ij, management server s
(i+1) kBe s
IjChild node and unique, s
(i+1) kHas only unique child s
(i+2) kAnd s
(i+2) kBe service server, with s
(i+1) kMerge to s
Ij
Wherein, s
IjRepresent i layer, j platform management server; In the merging process of management server, horizontal direction merges will have precedence over the vertical direction merging; The vertical direction that merges is bottom-up; Server after level merges is replaced by the many servers of former load.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210303088.5A CN102833333B (en) | 2012-08-23 | 2012-08-23 | Scalable network server group frame system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210303088.5A CN102833333B (en) | 2012-08-23 | 2012-08-23 | Scalable network server group frame system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102833333A true CN102833333A (en) | 2012-12-19 |
| CN102833333B CN102833333B (en) | 2015-01-28 |
Family
ID=47336300
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210303088.5A Active CN102833333B (en) | 2012-08-23 | 2012-08-23 | Scalable network server group frame system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102833333B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105681405A (en) * | 2016-01-06 | 2016-06-15 | 山东中创软件工程股份有限公司 | Network transmission monitoring method |
| CN106302726A (en) * | 2016-08-15 | 2017-01-04 | 浪潮(北京)电子信息产业有限公司 | The monitoring method of the management node of a kind of distributed type assemblies data and module |
| CN110928667A (en) * | 2019-12-06 | 2020-03-27 | 北京明略软件系统有限公司 | Task scheduling method and device, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5687319A (en) * | 1994-09-27 | 1997-11-11 | International Business Machines Corporation | Method and system for determining maximum cable segments between all possible node to node paths on a serial bus |
| CN101119364A (en) * | 2007-09-13 | 2008-02-06 | 上海大学 | Authenticating Ad Hoc group cipher key negotiation protocol |
| CN101420445A (en) * | 2007-10-25 | 2009-04-29 | 厦门大学 | Fast routing protocol of wireless sensor network |
-
2012
- 2012-08-23 CN CN201210303088.5A patent/CN102833333B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5687319A (en) * | 1994-09-27 | 1997-11-11 | International Business Machines Corporation | Method and system for determining maximum cable segments between all possible node to node paths on a serial bus |
| CN101119364A (en) * | 2007-09-13 | 2008-02-06 | 上海大学 | Authenticating Ad Hoc group cipher key negotiation protocol |
| CN101420445A (en) * | 2007-10-25 | 2009-04-29 | 厦门大学 | Fast routing protocol of wireless sensor network |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105681405A (en) * | 2016-01-06 | 2016-06-15 | 山东中创软件工程股份有限公司 | Network transmission monitoring method |
| CN106302726A (en) * | 2016-08-15 | 2017-01-04 | 浪潮(北京)电子信息产业有限公司 | The monitoring method of the management node of a kind of distributed type assemblies data and module |
| CN106302726B (en) * | 2016-08-15 | 2019-05-28 | 浪潮(北京)电子信息产业有限公司 | A kind of monitoring method and module of the management node of distributed type assemblies data |
| CN110928667A (en) * | 2019-12-06 | 2020-03-27 | 北京明略软件系统有限公司 | Task scheduling method and device, electronic equipment and storage medium |
| CN110928667B (en) * | 2019-12-06 | 2022-03-01 | 北京明略软件系统有限公司 | Task scheduling method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102833333B (en) | 2015-01-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103873279B (en) | Server management method and server management device | |
| CN104113606B (en) | The method of work of the distributed meta data node architecture of uniformity dynamic equalization | |
| CN103797463A (en) | Method and apparatus for assignment of virtual resources within a cloud environment | |
| CN105227364B (en) | A kind of method and device of ERP system resource-sharing | |
| CN105978950A (en) | Server resource distribution method, apparatus and system thereof | |
| CN102520933A (en) | Method and device for establishing tree menu based on user right | |
| CN201616710U (en) | Telescopic terminal concurrency access and load balancing system framework | |
| CN102035859A (en) | Method, system and device for displaying upload progress | |
| CN103428008A (en) | Big data distribution strategy oriented to multiple user groups | |
| CN101388997A (en) | Method and system for establishing video conference | |
| CN102833333A (en) | Scalable network server group frame system | |
| CN108347477A (en) | Data transmission method for uplink, device and server | |
| CN106664214A (en) | Failure recovery method and device for virtual network | |
| CN102625140A (en) | Method for frame-granularity real-time downloading and play of streaming media server | |
| CN106453576A (en) | Interactive method, interactive system and dispatching centre based on hybrid cloud platform | |
| CN103763206B (en) | Network scheduling method and gateway | |
| CN101467132B (en) | Method and system for distributing data processing units in a communication network | |
| CN104660690A (en) | Cloud video service monitoring system | |
| CN107346270A (en) | Method and system based on the sets cardinal calculated in real time | |
| CN107562547A (en) | A kind of CTDB group systems and creation method, create system | |
| CN201387555Y (en) | Comprehensive remote control monitoring system | |
| CN104935614B (en) | Data transmission method and device | |
| CN108170527B (en) | Remote multi-activity distributed message consumption method and device | |
| CN1905516B (en) | Method of allocating bidirection forwarding detecting conversational authentication code | |
| CN106453637B (en) | The method, apparatus and cloud platform of cloud platform efficient multiplexing server resource |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20201125 Address after: 510000 room 3, Tong Dong Road, Tianhe District, Guangzhou, Guangdong B-306 Patentee after: GUANGZHOU XUNHONG NETWORK TECHNOLOGY Co.,Ltd. Address before: 510640, 381, five mountain road, Guangzhou, Guangdong, Tianhe District, China Patentee before: SOUTH CHINA University OF TECHNOLOGY |
|
| TR01 | Transfer of patent right |