CN115208883B - Distributed system configuration generation method suitable for securities trading system - Google Patents
Distributed system configuration generation method suitable for securities trading system Download PDFInfo
- Publication number
- CN115208883B CN115208883B CN202210776976.2A CN202210776976A CN115208883B CN 115208883 B CN115208883 B CN 115208883B CN 202210776976 A CN202210776976 A CN 202210776976A CN 115208883 B CN115208883 B CN 115208883B
- Authority
- CN
- China
- Prior art keywords
- node
- cluster
- file
- configuration
- multicast
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/02—Details
- H04L12/16—Arrangements for providing special services to substations
- H04L12/18—Arrangements for providing special services to substations for broadcast or conference, e.g. multicast
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/06—Protocols specially adapted for file transfer, e.g. file transfer protocol [FTP]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/2866—Architectures; Arrangements
- H04L67/30—Profiles
Abstract
The invention relates to the technical field of securities trade, in particular to a distributed system configuration generation method applicable to a securities trade system, which comprises the following steps: constructing a source file for generating a configuration file; uploading the source file to a configuration center server; carrying out one-time generation test on the source file; if the verification is correct, storing the uploaded source file packet into a high-availability database, returning the version number to the request client, and if the verification is not correct, returning an error; sending a generation request to a server; the server generates a configuration file package and returns the configuration file package to the client. Compared with the prior art, the invention has the advantages that: the method only needs to manually declare the deployment and communication requirements of all parts of the transaction system, and the configuration file is generated by the program according to the preset rule, so that the probability of configuration file writing errors is greatly reduced, and the efficiency of configuration file generation is also improved; the local mode is set to provide flexibility, and the requirements of debugging and emergency are met.
Description
Technical Field
The invention relates to the technical field of securities trade, in particular to a distributed system configuration generation method suitable for a securities trade system.
Background
The steel stock and option trading system (hereinafter referred to as trading system) is a distributed system. A distributed system refers to a software system built on top of software. In a transaction system, the smallest logical component is a node, i.e., an individual program that can independently complete a set of logic, often in the form of a process. A plurality of nodes with the same logic function form a cluster, and the transaction system consists of a plurality of clusters.
To increase order processing efficiency, trading systems employ schemes that differ from conventional distributed systems. In a transaction system, nodes communicate in a multicast mode, each multicast channel needs to bind an IP and a port, and the information must be determined before the transaction system is started. In addition, in order to improve the operation efficiency of each node, the transaction system does not adopt a container arrangement mode to construct a cluster, but deploys the nodes to directly operate on a physical machine with only an operating system and necessary operation components.
To deploy the entire trading system, the deployment of each cluster must be specified, such as cluster name, number of nodes, etc. And to run a certain node, it must be deployed to a certain computer according to certain requirements. Meanwhile, a plurality of nodes may be distributed on the same or different computers, and normal communication between the nodes must be ensured. Therefore, we have to compile a set of profiles for the transaction system, deploy the transaction system according to some profiles, and let the nodes start and run through the profiles.
A set of configuration files that may be used to deploy a transaction system typically contains the following information:
(1) Infrastructure information. Indicating the underlying information such as physical machine information, network information, etc. that may be used to deploy the transaction system. Including host name, network card binding IP, etc.
(2) Deployment information. Indicating how each cluster and each node should be deployed, including the number of nodes that the cluster contains, the name of the host that each node is to deploy, the location where the node is running on the host, etc.
(3) And (5) communication information. The communication information, namely multicast information in the transaction system, is specifically divided into three parts: firstly, at the node level, each node in the same cluster needs to bind a multicast channel to complete intra-cluster communication, and the channel is called intra-cluster channel; secondly, at the node type level, all clusters under certain node types need to issue the same type of service information outwards respectively, so that a multicast channel is bound to each cluster, and the channels are called service channels. Thirdly, at the node type level, between certain two node types, the cluster needs to establish multicast channels corresponding to each other, and the channels have only one sender and also only one receiver, which are called proprietary channels. The combination of the IP and port numbers bound by the three multicast channels is not allowed to be repeated, and each multicast channel has a globally unique multicast number.
It is noted that, in order to facilitate multiplexing by multiple persons, some fine adjustment of port numbers in the multicast information allocated according to the above rule is required. The method is characterized in that each user group is allocated with a two-digit number called an environment number, and the environment number is replaced by a certain two digits of a port number, so that a plurality of users can deploy and operate the same set of configuration schemes on the same group of physical machines at the same time.
(4) Node configuration files. I.e. the configuration files required by a certain class of nodes when running. These files are related to business logic, and the node needs to read the configuration file corresponding to the type to which the node belongs when running.
For a transaction system, its stability is directly linked to the property security of people and the financial security of countries. Therefore, it is important to ensure the stability of the transaction system. Unlike the dynamic management of conventional distributed systems, the transaction system adopts static configuration, static deployment and static operation modes without considering faults, so as to ensure that uncontrollable variables are not introduced during the operation of the system.
In special cases, the transaction system also has the possibility to temporarily change the configuration and deployment. Meanwhile, in order to facilitate testing, a developer needs to have a method for quickly adjusting the configuration of the transaction system. The profile generation method of the transaction system must take this into account.
In the prior operation practice, when the system operates normally, operation staff firstly confirms deployment information of a transaction system, such as cluster distribution, node quantity and the like, manually distributes multicast channels for communication for the clusters and the nodes according to an infrastructure list, and then writes a set of configuration files which can finally support program operation for each node according to the deployment information, the multicast channels and configuration files related to the node types.
In special cases, the operator may enable another set of configuration files that are preset to alter the transaction system. There may be multiple sets of such configuration files to cope with different scenarios, where the writing process of any set is consistent with the writing flow when the system is operating normally.
The prior configuration file writing mode has some defects. The writing efficiency is low, careless mistakes are easy to occur, the debugging is inconvenient, and the workload of operation and maintenance personnel is increased. Errors generated during the writing of the configuration files are often difficult to perceive and locate in subsequent system tests, which often compromises the development efficiency of the whole system.
Furthermore, the multicast information of the whole transaction system is manually written, and the workload thereof is remarkable. We can assume that there are M node types in the transaction system, each node type includes N clusters, each cluster includes X nodes, and further assume that each node type establishes a proprietary channel with a node type other than itself. For ease of statistics, we record a piece of multicast information for each node corresponding to a certain multicast channel as a unit.
In the case that only the publisher is considered, the number of multicast channels that each cluster should establish for another node type is N, the multicast information that each cluster newly adds for another node type is n×x units, the multicast information that each cluster belongs to for another node type newly adds for n×x×n units, and then the multicast information units that each cluster newly adds for M node types are:
M*(M-1)*N*X*N
as can be seen from this, the number of multicast messages increases sharply with the expansion of the system, because only the multicast messages of the publisher are proportional to the number of secondary nodes, the number of clusters, and the number of nodes included in the system. In a transaction system, the three are generally at a high level, and the workload of manually writing multicast information is very large.
If the multicast information is manually written, the problems of low efficiency, easy error, difficult inspection and the like in the writing process are amplified along with the expansion of the transaction system, and finally pressure is brought to the stable operation of the whole transaction system.
The prior art scheme is a transaction system configuration management and generation method which is totally dependent on manual execution, and the specific execution steps are as follows:
step one: according to a predetermined deployment scheme, determining deployment information, wherein the method specifically comprises the steps of determining the node type, cluster name, cluster number, node number in each cluster, node deployment information and the like of a transaction system to be started. Each cluster is assigned a globally unique cluster number, and each node is assigned a globally unique node number. In order to construct a multicast channel, a relative offset of each node within its cluster, referred to as a node index, needs to be allocated, and a relative offset of each cluster in the node type to which it belongs, referred to as a cluster index, needs to be allocated. The hierarchical structure of the transaction system is shown in fig. 1.
Step two: according to a preset communication scheme, three types of multicast channels, namely one-to-one multicast channels among nodes in a cluster, one-to-one multicast channels among clusters and one-to-many multicast channels among clusters, are constructed, wherein each channel comprises a globally unique multicast number and simultaneously comprises a globally unique combination of an IP address and a port number.
The specific flow of constructing the communication information is as follows:
2.1. selecting a cluster which does not determine communication information in a transaction system, and setting the cluster as a cluster A;
2.2. let node type A include A 1 、A 2 Two clusters, A 1 Comprises 2 nodes a 1 、a 2 ，A 2 Comprises 2 nodes a 3 、a 4 ，
2.3. For cluster A 1 Two nodes a of (a) 1 、a 2 Distributing multicast channel t in cluster 1 、t 2 . In order for the nodes to be able to resolve the communication information, the multicast channels between the nodes in the cluster need to be additionally bound with the corresponding node numbers and node indexes. t is t 1 、t 2 Related information at A 1 Recording is carried out in all nodes of the cluster;
2.4. selecting a cluster which is not distributed with multicast channels in the cluster in the node type A, and repeating the step 2.3 until all clusters are distributed completely;
2.5. and allocating service channels. Let node type a pre-specify a need for a pair of multicast communications with node type B, which includes cluster B 1 、B 2 . Each cluster in a must bind a multicast channel as the sender. Here multicast channel t is allocated 3 、t 4 Respectively for cluster A 1 、A 2 Is a pair of multicast communications.
In order for the communication partners to be able to resolve the communication information, the multicast channel here needs to bind the cluster number and cluster index of the sender cluster.
t 3 Is in cluster A 1 、B 1 、B 2 All nodes are recorded, t 4 Is in cluster A 2 、B 1 、B 2 All the nodes are recorded;
2.6. a proprietary channel is allocated. Let node type a pre-designate to perform one-to-one multicast communication with node type C, which contains cluster C 1 、C 2 Multicast channels need to be allocated to each cluster in a according to the number of clusters in C. Here, t is allocated 5 、t 6 For A 1 And C 1 、C 2 Is allocated t 7 、t 8 For A 2 And C 1 、C 2 Is a communication of (a).
In order for the communication parties to be able to resolve the communication information, the multicast channel here needs to bind the cluster number and cluster index of the sender cluster and the cluster number and cluster index of the receiver cluster at the same time.
t 5 Is in cluster A 1 、C 1 All the nodes included are to be recorded,t 6 is in cluster A 1 、C 2 All nodes are recorded, t 7 Is in cluster A 2 、C 1 All nodes are recorded, t 8 Is in cluster A 2 、C 2 All the nodes are recorded;
step three: and respectively compiling configuration files of each node according to the deployment information obtained in the first step and the communication information obtained in the second step. Generally, each node type corresponds to a set of configuration files with the same structure, but when the configuration files are deployed on a specific physical machine, deployment information and communication information on the configuration files need to be modified into information in actual operation.
The prior art scheme mainly has the following defects:
1. the manual distribution and writing mode is adopted, so that the efficiency is low, and the error rate is high. In the existing configuration file writing mode, deployment information and communication information are generated manually. As the number of nodes increases, the number of communication information increases with the power correlation, and errors are likely to occur during manual processing. Errors introduced by the configuration file are typically hidden and can have a significant impact on the efficiency of the system when it is being debugged.
2. There is no highly available and reliable way of storing configuration files. The input information used for generating the configuration file, such as infrastructure information, communication statement, configuration file of node type, etc., are stored in some computers in a centralized or decentralized manner, so that there is a risk of losing or damaging, and centralized management is inconvenient.
3. There is a repetitive effort. When the node configuration file is compiled after the deployment information and the communication information are determined, the same set of templates are generally used for compiling the same node type, and only the entries related to deployment and communication are replaced. This process involves some repetitive labor and is prone to human error.
4. Verification is lacking. The manually filled configuration file cannot be verified and can only rely on system testing.
5. Not flexible enough. In special cases, the configuration scheme can only be changed by means of a preset configuration scheme which is tested, and the flexibility is still lacking.
Disclosure of Invention
The invention aims to solve the defects of the prior art and solve a plurality of problems in the writing process of configuration files, and designs a set of transaction system configuration management and generation method based on Golang template filling technology from the aspect of information layering.
To achieve the above object, there is provided a distributed system configuration generation method suitable for a securities trading system, the method comprising:
s1, constructing a source file for generating a configuration file, wherein the source file comprises an infrastructure information file, deployment information, a service statement file and a configuration file template.
S2, using the configuration center client to upload all configuration schemes and package the configuration schemes into a source file to the configuration center server.
And S3, immediately performing one-time generation test on the source file after the server receives the source file.
S31, generating deployment information, selecting an unverified scheme P, selecting a cluster C1 which does not process the deployment information in the scheme P, distributing a cluster number for the cluster C1, and calculating a cluster index of the cluster C in the type of the node; reading a deployment information file of the cluster C1, selecting an unprocessed node, and distributing a node number and a node index for the unprocessed node; a multicast channel information table is newly built for each cluster in scheme P and assumes the available IP range, available port numbers and assumed environment numbers.
S32, generating communication information, and selecting a node type T of which the communication information is not processed in the scheme P 1 Reading a service statement file thereof, reading a intra-cluster channel part in the service statement file, and recording a node type T 1 Network name W that internal communication needs to use 1 Select node type T 1 Cluster C in which no intra-cluster multicast channel has been generated 2 Selected cluster C 2 Not yet processed node N 1 According to the network name W 1 Query node N 1 Is a host of (a)Network card information, then to N 1 Distributing multicast channels in a cluster, replacing the designated position of the original port number by using an environment number in the process of generating multicast channel information, ensuring that the replaced port number is globally unique, trying out the next original port number again if repetition is found, and returning an error if the original port number is used up; to cluster C 2 Write-in node N 1 Multicast information in the cluster; selected node type T 1 A network name W which is not yet processed 2 The method comprises the steps of carrying out a first treatment on the surface of the Select network name W 2 The next service type multicast statement, i.e. a one-to-many multicast statement among clusters, is noted as statement B 1 The method comprises the steps of carrying out a first treatment on the surface of the Select T 1 Statement B not yet processed 1 Cluster C of (2) 3 The method comprises the steps of carrying out a first treatment on the surface of the Query C 3 The corresponding cluster number and cluster index is cluster C 3 The allocation corresponds to B 1 Service type multicast channel R of (1) 1 And multicast number U 1 The method comprises the steps of carrying out a first treatment on the surface of the Selected cluster C 3 Not yet processed node N 2 Searching for deployment node N 2 The host corresponds to the network name W 2 Network card information, node number and node index of the network, and multicast channel R 1 Multicast number U 1 The cluster C is recorded with the sender cluster number, the sender cluster index, the node number, the node index and the network card information 3 In (a) and (b); selected node type T 1 Mid-network name W 2 A proprietary type multicast statement under the network, i.e. a one-to-one type multicast statement among clusters, which is noted as B 2 Its receiver node type is T 2 The method comprises the steps of carrying out a first treatment on the surface of the Selected node type T 1 One of the outstanding claims B 2 Cluster C of (2) 4 Query cluster C 4 Corresponding cluster number M 1 And cluster index I 1 The method comprises the steps of carrying out a first treatment on the surface of the Selected node type T 2 One of the outstanding claims B 2 Cluster C of (2) 5 Query cluster C 4 Corresponding cluster number M 2 And cluster index I 2 Then cluster C 4 And cluster C 5 Allocation corresponds to claim B 2 Dedicated type multicast channel R of (c) 2 And multicast number U 2 The method comprises the steps of carrying out a first treatment on the surface of the Selected transmission Fang Jiqun C 4 One of the outstanding claims B 2 Node N of (2) 3 Searching for deployment node N 3 The host corresponds to the network name W 2 Network card information, node number and node index of the network, and multicast channel R 1 Multicast number U 1 Sender Cluster number M 1 And cluster index I 1 Group number M of receiver 2 And cluster index I 2 Node N 3 Node number, node N of (a) 3 Node index of (1), network card information record cluster C 4 In (a) and (b); selecting a recipient Cluster C 5 One of the outstanding claims B 2 Node N of (2) 4 Searching for deployment node N 4 The host corresponds to the network name W 2 Network card information, node number and node index of the network, and multicast channel R 1 Multicast number U 1 Sender Cluster number M 1 And cluster index I 1 Group number M of receiver 2 And cluster index I 2 Node N 4 Node number, node N of (a) 4 Node index of (1), network card information record cluster C 5 Is a kind of medium.
S33, filling configuration templates, and selecting a node type T of which the configuration templates are not filled in P 3 The method comprises the steps of carrying out a first treatment on the surface of the Selected node type T 3 Cluster C of which no configuration file has been generated 6 The method comprises the steps of carrying out a first treatment on the surface of the Selected cluster C 6 Node N of which no configuration file has been generated 5 The method comprises the steps of carrying out a first treatment on the surface of the Opposite node N 5 Constructing a temporary file F usable for populating a configuration template 1 Which includes node N 5 Node number, node index, cluster C 6 Cluster number, cluster index, cluster name; filling node type T by Golang template filling technique using temporary files 3 All the configuration templates below, a group of configuration templates suitable for the node N is obtained 5 Is a configuration file of (a);
and S4, if the verification in the step 3 is correct, storing the uploaded source file packet into a high-availability database, distributing a globally unique version number for the source file packet, returning the version number to the request client, recording the version number into the high-availability database, and if the verification is not correct, returning an error.
S5, when the client requests to generate the configuration file, firstly querying the version number of the source file package existing in the server, selecting a target version number, and sending a generation request to the server.
S6, the server generates a configuration file packet according to the step 3 and returns the configuration file packet to the client.
Collecting information of each available physical machine, including a host name, all network names, network card names positioned in different networks, IPv4 addresses and IPv6 addresses bound by each network card, the position of the host and the like, and writing the information into a file by using any text serialization format; the construction and deployment information specifically comprises the following steps: constructing a deployment information file by a cluster level, wherein the number of nodes of the cluster and the name of a host computer to be deployed by each node are required to be indicated in the file; the construction of the service statement file is specifically as follows: constructing a service declaration file at the node type level for indicating the type of communication information, the file comprising three parts, one being intra-cluster channels, the part having to indicate on which network the node type communicates; secondly, the service channel, the part must specify the service channel to be published and the name of the service channel to be subscribed to; thirdly, a proprietary channel, wherein the part must specify the node type of a receiver of the proprietary channel; the construction of the configuration file template specifically comprises the following steps: and constructing a configuration file template at the node type level, generating actual configuration files of all nodes under the node type by virtue of Golang template filling technology, wherein available replacement items on the configuration file template comprise cluster names, cluster numbers, cluster indexes, node numbers and node indexes, and filling positions requiring replacement on the template by using actual deployment information of the node when generating the configuration file of a certain node.
Preferably, a plurality of sets of conventional configuration schemes can be prepared at the same time according to actual needs, and the step S1 is repeated until the setting of all the configuration schemes is completed, and the schemes are placed in the same configuration file package.
Preferably: the step S31 also includes newly creating a deployment information table depth list. Json, cluster C 1 Cluster name, cluster number, clusterAfter the unprocessed node is selected, the corresponding host information is found according to the host name of the unprocessed node on the infrastructure information file index.
Preferably, in the step S31, all nodes in all clusters in the cyclic traversal scheme P are distributed and recorded.
Preferably, the step S32 is performed circularly until the node type T 1 All the clusters in the network complete the generation of multicast channels in the clusters and the node type T 1 All clusters in (1) have been assigned claim B 1 Corresponding multicast channel and node type T 1 All network names W in 2 The following service statement is distributed with multicast channel and cluster C 4 All nodes in (1) have recorded declaration B 2 Multicast information of (C) cluster 5 All nodes in (1) have recorded declaration B 2 Multicast information up to sender cluster C 4 Corresponding to proprietary claim B with all receiver clusters 2 Is distributed and recorded until the node type T 1 All clusters in (1) correspond to claim B 2 Is distributed and recorded until the node type T 1 All network names W in 2 Proprietary declarations under the network have all been assigned multicast channels up to node type T 1 The service type declaration and the proprietary declaration under all networks are distributed and recorded completely until the communication information of all node types in the scheme P is processed and recorded completely.
The step S33 loops until the cluster C 6 All nodes in the network generate corresponding configuration files until the node type T 3 Corresponding configuration files are generated by all nodes in the scheme P until the corresponding configuration files are generated by all nodes in the scheme P, until all configuration schemes in the source file package pass verification, and if errors occur, the errors are returned.
Compared with the prior art, the invention has the advantages that:
1. the method only needs to manually declare the deployment and communication requirements of all parts of the transaction system, and the configuration file is generated by the program according to the preset rule, so that the probability of configuration file writing errors is greatly reduced, and the efficiency of configuration file generation is also improved.
2. The high-availability storage cluster is used for storing the source files for generating the configuration files, so that the risk of losing or damaging the configuration files is greatly reduced, and the centralized management of the configuration files is realized through the configuration center server.
3. In the process of generating the configuration file, the program automatically checks, and if errors on some corresponding relations occur in the configuration file, for example, deployment information of the transaction system is not matched with infrastructure information, some clusters of the transaction system monitor an unpublished multicast, and the like, the program will report errors in the process of generating and filling the configuration file.
4. The local mode is set to provide flexibility, and the requirements of debugging and emergency are met.
5. The method has the advantages that the version management is carried out on the infrastructure information data and the configuration file package, so that the staff can conveniently and quickly find and position the configuration file of the whole transaction system, and the maintenance and management are also facilitated.
Drawings
FIG. 1 is a schematic diagram of a transaction system hierarchy of the present invention;
FIG. 2 is a schematic diagram of a configuration center assembly of the present invention;
FIG. 3 is an on-line mode flow chart of the present invention;
fig. 4 is a flow chart of a local mode of the invention.
Detailed Description
Referring to fig. 1, 2, 3 and 4, a distributed system configuration generation method suitable for a securities trading system, specifically comprises the following steps:
1. (step S1) constructing a source file for generating a configuration file
1.1. An infrastructure information file is constructed. And collecting information of each available physical machine, including a host name, all network names, network card names positioned on different networks, IPv4 addresses and IPv6 addresses bound by each network card, the position of the host and the like, and writing the information into a file by using any text serialization format. Assuming that json format is used here, the written file is infracuctive.
1.2. And constructing deployment information. A deployment information file is built at the cluster level, in which the number of nodes of the cluster and the name of the host that each node needs to deploy should be specified. Assuming json format serialization is used here, each deployment information file is named depth.
1.3. A service declaration file is constructed. Constructing a service declaration file at the node type level for indicating the type of communication information, the file comprising three parts, one being intra-cluster channels, the part having to indicate on which network the node type communicates; secondly, the service channel, the part must specify the service channel to be published and the name of the service channel to be subscribed to; third, the proprietary channel, which must specify the node type of the proprietary channel receiver. Each service announcement file is named service.
1.4. And constructing a configuration file template. And constructing a configuration file template at the node type level, and generating actual configuration files of all nodes under the node type by using a Golang template filling technology. Alternatives available on the profile template include cluster name, cluster number, cluster index, node number, node index. When a configuration file of a certain node is generated, the actual deployment information of the node is used to fill in the position requiring replacement on the template.
Thus, a set of configuration schemes is completed. And (3) preparing a plurality of sets of conventional configuration schemes according to actual needs, and repeating the steps 1.2-1.4 until the setting of all the configuration schemes is completed. These schemes are placed in the same profile package.
2. (step S2) using the configuration center client to upload all configuration schemes and packages into a source file to the configuration center server. If the first upload is performed, the infrastructure information file also needs to be uploaded separately.
3. And (S3) immediately performing a generation test on the source file once after the server receives the source file. The specific flow of the test is as follows, wherein step 3.1-step 3.6 generates deployment information (step S31), step 3-7-step 3.31 (step S32) generates communication information, and step 3.32-step 3.39 (step S33) is to fill the configuration template:
3.1. An unverified scheme P is selected. Newly creating a deployment information table depth list.
3.2. Selecting cluster C in P which does not process deployment information 1 It is assigned a cluster number and its cluster index in the type of the belonging node is calculated. The cluster name, cluster number, and cluster index are written into the deleymentlist.
3.3. Read C 1 Selecting an unprocessed node, allocating node numbers and node indexes for the unprocessed node, and finding corresponding host information on the infrastructure information file index. The node number, node index, and host information of the node are recorded into a deleymentlist. If the host name does not appear on the infrastructure information file, reporting an error and returning.
3.4. Repeating step 3.3 until C 1 And (3) all nodes are distributed and recorded.
3.5. Repeating the steps 3.2-3.4 until all clusters in P are distributed and recorded.
3.6. A multicast channel information table topicinfo. Json is newly built for each cluster in the scheme P, and the available IP range is assumed to be any IP, the available port number is 1024-65535, and the environment number is assumed to be 00.
3.7. A node type T in the selected scheme P for which no communication has been processed 1 And reading the service statement file service.
3.8. Reading the intra-cluster channel part in service. Json, and recording T 1 Network name W that type internal communication needs to use 1 。
3.9. Select T 1 Cluster C in which no intra-cluster multicast channel has been generated 2 。
3.10. Select C 2 Not yet processed node N 1 . In depoymentIn List.json, according to W 1 Query N 1 Is provided. Then to N 1 Multicast channels within the cluster are allocated.
In the process of generating the multicast channel information, the designated position of the original port number needs to be replaced by using the environment number, the global uniqueness of the replaced port number is ensured, if repetition is found, the next original port number is tried again, and if the original port number is used up, an error is returned.
To C 2 N is written in TopicInfo.json 1 Is provided, comprising: sender cluster name, sender cluster number, sender cluster index, multicast channel information, multicast number, node number, and host network card information.
3.11. Repeating step 3.10 until C 2 The nodes in the network are all processed.
3.12. Repeating steps 3.9-3.11 until T 1 All the clusters in the network are used for completing the generation of multicast channels in the clusters.
3.13. Select T 1 A network name W which is not yet processed 2 (W 2 And W is equal to 1 Repeatable).
3.14. Selecting W 2 The next service type multicast statement, i.e. one-to-many multicast statement among clusters, is denoted as B 1 。
3.15. Select T 1 One of the outstanding claims B 1 Cluster C of (2) 3 In depth list. Json, query C 3 Corresponding cluster numbers and cluster indexes. Then C is 3 The allocation corresponds to B 1 Service type multicast channel R of (1) 1 And multicast number U 1 。
3.16. Select C 3 Not yet processed node N 2 . And find deployment N in the depoymentList. Json 2 The host corresponds to W 2 Network card information, node number and node index of the network. Multicast channel R 1 Multicast number U 1 Record C is set up to the sender cluster number, sender cluster index, node number, node index and network card information 3 In topicinfo. Json.
3.17. Repeating step 3.16 until C 3 All nodes in the network are recorded.
3.18. Repeating the steps 3.15-3.17 until T 1 All clusters in (1) have been assigned B 1 Multicast channels corresponding to the service declarations.
3.19. Repeating steps 3.14-3.18 until T 1 All W of (3) 2 The service claims under the network are all allocated multicast channels.
3.20. Select T 1 Middle W 2 A proprietary type multicast statement under the network, i.e. a one-to-one type multicast statement among clusters, which is noted as B 2 Its receiver node type is T 2 。
3.21. Select T 1 One of the outstanding claims B 2 Cluster C of (2) 4 In depth list. Json, query C 4 Corresponding cluster number M 1 And cluster index I 1 。
3.22. Select T 2 One of the outstanding claims B 2 Cluster C of (2) 5 In depth list. Json, query C 4 Corresponding cluster number M 2 And cluster index I 2 . Then C is 4 And C 5 The allocation corresponds to B 2 Dedicated type multicast channel R of (c) 2 And multicast number U 2 。
3.23. Selected transmission Fang Jiqun C 4 One of the outstanding claims B 2 Node N of (2) 3 And find deployment N in the depoymentList. Json 3 The host corresponds to W 2 Network card information, node number and node index of the network. Multicast channel R 1 Multicast number U 1 Sender Cluster number M 1 And cluster index I 1 Group number M of receiver 2 And cluster index I 2 、N 3 Node number, N of (2) 3 Node index of (1), network card information record C 4 In topicinfo. Json.
3.24. Repeating step 3.23 until cluster C 4 All nodes in (1) have recorded declaration B 2 Is provided).
3.25.Selecting a recipient Cluster C 5 One of the outstanding claims B 2 Node N of (2) 4 And find deployment N in the depoymentList. Json 4 The host corresponds to W 2 Network card information, node number and node index of the network. Multicast channel R 1 Multicast number U 1 Sender Cluster number M 1 And cluster index I 1 Group number M of receiver 2 And cluster index I 2 、N 4 Node number, N of (2) 4 Node index of (1), network card information record C 5 In topicinfo. Json.
3.26. Repeating step 3.25 until cluster C 5 All nodes in (1) have recorded declaration B 2 Is provided).
3.27. Repeating steps 3.22-3.26 until the sender cluster C 4 Corresponding to proprietary claim B with all receiver clusters 2 And (3) completing the distribution and recording of the multicast information.
3.28. Repeating steps 3.21-3.27 until node type T 1 All clusters in (1) correspond to claim B 2 And (3) completing the distribution and recording of the multicast information.
3.29. Repeating steps 3.20-3.28 until T 1 All W of (3) 2 Proprietary declarations under the network have been assigned multicast channels.
3.30. Repeating steps 3.13-3.29 until T 1 And (3) distributing and recording all the business type declarations and proprietary declarations under the network.
3.31. And repeating the steps 3.7-3.30 until all the node type communication information in the scheme P is processed and recorded.
3.32. Selecting a node type T of P which is not filled with configuration templates 3 。
3.33. Select T 3 Cluster C of which no configuration file has been generated 6
3.34. Select C 6 Node N of which no configuration file has been generated 5 。
3.35. Opposite node N 5 Constructing a temporary file F usable for populating a configuration template 1 Wherein comprises N 5 Node number, node index, C 6 Cluster number, cluster index, cluster name.
3.36. Using temporary files F 1 Filling node type T by means of Golang template filling technology 3 All the configuration templates below, a group of configuration templates suitable for the node N is obtained 5 Is a configuration file of (a).
3.37. Repeating the steps 3.34-3.36 until the cluster C 6 Corresponding configuration files are generated by all nodes in the network.
3.38. Repeating the steps 3.33-3.37 until the node type T 3 Corresponding configuration files are generated by all nodes in the network.
3.39. Repeating the steps 3.32-3.38 until all nodes in the scheme P have generated corresponding configuration files.
3.40. Repeating the steps 3.1-3.39 until all configuration schemes in the source file package pass verification. If an error occurs, an error is returned.
4. (step S5) if the check in the step 3 (step S3) is correct, storing the uploaded source file package into a high-availability database, distributing a globally unique version number for the source file package, returning the version number to the request client, and recording the version number into the high-availability database. If the verification is not passed, an error is returned.
5. When the client requests to generate the configuration file, firstly, inquiring the version number of the source file package existing in the server, selecting a target version number, and sending a generation request to the server.
6. (step S6) the server generates a configuration file package according to the step 3 and returns the configuration file package to the client.
In the local mode, the flow chart of the method is shown in fig. 4, and the specific flow is as follows:
1. the source file package is built locally, and the steps are the same as step 1 of the normal mode.
2. And opening a configuration center client, selecting a local mode completely dependent on a local source file, inputting a source file package path, outputting the path, and submitting a generation request.
3. The client generates a configuration file package in the same step as step 3 of the regular mode. If errors occur in the generation process, reporting errors, otherwise, placing the generated configuration file package under a designated output path.
Example 1
An implementation procedure of a distributed system configuration generation method applicable to a securities trading system according to the present invention is illustrated:
assuming that there are 4 physical machines in the system, host1 to host4 respectively, their relevant attributes are given in tables 1 to 4 respectively:
TABLE 1 physical machine host1 information table
TABLE 2 physical machine host2 information table
| Attributes of | Value of | Description of the invention |
| Host name | host2 | |
| Position of | jinqiao | |
| Network card 1 name | a4 | The network card is connected with the network intranet |
| Network card 1 | ipv4 address | 197.3.2.2 |
| Network card 2 name | a5 | The network card is connected with a network extranet |
| Network card 2 | ipv4 address | 10.112.3.2 |
| Network card 3 name | a6 | The network card is connected with a network oplan |
| Network card 3 | ipv4 address | 224.1.2.2 |
| User 1 name | cfgu | |
| User 2 name | xwang3 |
TABLE 3 physical machine host3 information table
TABLE 4 physical machine host4 information table
The infrastructure information file infrastrure. Json may be written as follows:
/>
/>
/>
assume that three node types currently need to be deployed on two platforms together, respectively:
mc: deployed on the DTP platform, two clusters are provided, each cluster comprising two nodes, wherein the two nodes of set1 need to be deployed on host1 and host2, respectively, and the two nodes of set2 need to be deployed on host3 and host4, respectively. The deployment paths of the two clusters are respectively/User/sseadmin/MC/set 1,/User/sseadmin/MC/set2
Cs: deployed on the DTP platform, and a common cluster, named set1, includes two nodes, and needs to be deployed on host1 and host2 respectively. Deployment path is/User/sseadmin/CS/set 1
RC: the two nodes of the set1 need to be deployed on host1 and host2 respectively, and the two nodes of the set2 need to be deployed on host3 and host4 respectively. The deployment paths of the two clusters are respectively/User/sseadmin/RC/set 1,/User/sseadmin/RC/set2
According to the construction rules, a depth is created for MC.set1. Json as follows:
the new depth for mc.set2. Json is as follows:
/>
similarly, the deployment information files of other clusters are built, and are not described herein.
Assume that the communication conditions of the system are as follows:
cs: on extranet, the setTopic type multicast channel is sent to MC and RC respectively. According to the construction rule, the names are DTP.MC and MTP.RC respectively; not actively subscribing; intra-cluster communication network using inter
RC: on the extranet, a bizTopic type multicast channel named risk is issued outwards; not actively subscribing; intra-cluster communication network using inter
Mc: not issuing; subscribing to bizTopic named risk on extranet; intra-cluster communication network using inter
Assuming that the multicast information sent by CS to dtp.mc is proposed by one of the engines, named engine1, and the multicast information sent to mtp.rc is proposed by the node type, an additional engine1 folder may be created under its service folder, and a service. Json file may be created for the CS alone.
According to the construction rule, a service declaration file service. Json is constructed for CS as follows
/>
The service declaration file service. Json established for engine1 is as follows:
The service declaration file service. Json is built for MC as follows
Constructing a service declaration file service. Json for an RC is as follows
Assuming that CS, RC, MC all use one configuration template app_shl.toml, the following is:
[Information]
platName＝"{{.PlatName}}"
nodeType＝"{{.NodeType}}"
nodeId＝{{.NodeId}}
nodeIndex＝{{.NodeIndex}}
setID＝{{.SetId}}
setName＝"{{.SetName}}"
setIndex＝{{.SetIndex}}
assuming that a unique configuration needs to be used for the so1 under the CS, a new so1 folder can be created under the template folder of the CS, and the configuration is added, assuming that its name is so1.Toml, as follows:
[Information]
platName＝"{{.PlatName}}"
in this example, it is assumed that the target version number to be submitted is 1.2.0, which contains one configuration scheme schema 1, under which the platform DTP and MTP are contained. The platform DTP contains node types CS and MC, the CS contains one cluster set1, and the MC contains 2 clusters set1 and set2. The platform MTP contains a node type RC under it, two clusters set1 and set2. The corresponding configuration folder should be constructed as follows:
at this time, a folder in which a source file is placed is selected, and a generation request is submitted, so that a generation result can be obtained at a target position, and the directory is:
/>
/>
let us take dtp.cs.set1 as an example, where the deployment information of the populated cluster is as follows:
/>
the generated multicast information file topicinfo. Json is as follows:
/>
/>
/>
host1_10 indicates that the node is deployed on host1, the node number is 10, and the content of the filled app_shell.toml file is as follows:
[Information]
platName＝"DTP"
nodeType＝"CS"
nodeId＝10
nodeIndex＝1
setID＝6
setName＝"set1"
setIndex＝2
The contents of the filled so1.Toml file are as follows:
[Information]
platName＝"DTP"
the rest of the files are similar to the above file structure, and only have changes in respective deployment information and specific values, and are not described here again.
The above description is only specific to the embodiments of the invention, but the scope of the invention is not limited thereto, and any person skilled in the art who is skilled in the art to which the invention pertains shall apply to the technical solution and the novel concept according to the invention, and shall all be covered by the scope of the invention.
Claims (8)
1. A distributed system configuration generation method for a securities trading system, the method comprising
The steps are as follows: after receiving the source file, the server side immediately performs a one-time generation test on the source file,
(1) Generating deployment information, selecting an unverified scheme P, and selecting clusters in the scheme P which do not process the deployment informationC 1 Assigning a cluster number to the node and calculating a cluster index of the node in the type of the node; read clustersC 1 Selecting an unprocessed node, and distributing node numbers and node indexes for the unprocessed node; a multicast channel information table is newly built for each cluster in the scheme P, and an available IP range, an available port number and an assumed environment number are assumed;
(2) Generating communication information, selecting a node type in scheme P for which communication information has not been processedT 1 Reading a service statement file thereof, reading a intra-cluster channel part in the service statement file, and recording node typesT 1 Network name that internal communication needs to useW 1 Selecting a node typeT 1 A cluster of multicast channels in the cluster not yet generatedC 2 Selected clustersC 2 Not yet processed a nodeN 1 According to the network nameW 1 Query nodeN 1 Is then directed toN 1 Distributing multicast channels in a cluster, replacing the designated position of a native port number by using an environment number in the process of generating multicast channel information, ensuring that the replaced port number is globally unique, trying out the next native port number again if repetition is found, returning an error if the native port number is used up, and returning to the clusterC 2 Write-in nodeN 1 Multicast information in the cluster; selecting a node typeT 1 A network name that has not been processedW 2 Selecting a network nameW 2 The next service type multicast statement, i.e. one-to-many multicast statement among clusters, is recorded as the statementB 1 The method comprises the steps of carrying out a first treatment on the surface of the SelectingT 1 A statement that has not been processedB 1 Is a cluster of (a)C 3 The method comprises the steps of carrying out a first treatment on the surface of the QueryingC 3 Corresponding cluster number and cluster index, which are clusters C 3 The allocation corresponds toB 1 Is a service type multicast channelR 1 And multicast numberU 1 The method comprises the steps of carrying out a first treatment on the surface of the Selecting clustersC 3 Not yet processed a nodeN 2 Searching for deployment nodesN 2 Corresponding to the network nameW 2 Network card information, node number and node index of the network, multicast channelR 1 Multicast numberU 1 The sender cluster number, the sender cluster index, the node number, the node index and the network card information are recorded into the clusterC 3 In (a) and (b); selecting a node typeMedium network nameW 2 A proprietary type multicast statement under the network, i.e. a one-to-one type multicast statement among clusters, which is noted asB 2 Its receiver node type isT 2 The method comprises the steps of carrying out a first treatment on the surface of the Select node type +.>One of the outstanding claimsB 2 Is a cluster of (a)C 4 Query clustersC 4 Corresponding cluster numberM 1 Cluster indexI 1 The method comprises the steps of carrying out a first treatment on the surface of the Selecting a node typeT 2 One of the outstanding claimsB 2 Is a cluster of (a)C 5 Query clustersC 4 Corresponding cluster numberM 2 Cluster indexI 2 Then is a clusterC 4 And clustersC 5 Assigning declarationsB 2 Is a proprietary type multicast channelR 2 And multicast numberU 2 The method comprises the steps of carrying out a first treatment on the surface of the Selected transmission Fang JiqunC 4 One of the outstanding claimsB 2 Is a node of (2)N 3 Searching for deployment nodesN 3 Corresponding to the network nameW 2 Network card information, node number and node index of the network, multicast channelR 1 Multicast number U 1 Transmitting Fang JiqunM 1 Cluster indexI 1 Group number of receiverM 2 Cluster indexI 2 NodeN 3 Node number, node of (a)N 3 Node index of (1), network card information record clusterC 4 In (a) and (b); selecting a group of recipientsC 5 One of the outstanding claimsB 2 Is a node of (2)N 4 Searching for deployment nodesN 4 Corresponding to the network nameW 2 Network card information and node number of networkAnd node index, multicast channelR 1 Multicast numberU 1 Transmitting Fang JiqunM 1 Cluster indexI 1 Group number of receiverM 2 Cluster indexI 2 NodeN 4 Node number, node of (a)N 4 Node index of (1), network card information record clusterC 5 In (a) and (b);
(3) Filling configuration templates, and selecting a node type of one of P which is not filled with the configuration templatesT 3 The method comprises the steps of carrying out a first treatment on the surface of the Selecting a node typeT 3 A cluster in which no configuration file has been generatedC 6 The method comprises the steps of carrying out a first treatment on the surface of the Selecting clustersC 6 A node in which no configuration file has been generatedN 5 The method comprises the steps of carrying out a first treatment on the surface of the Opposite nodeN 5 Constructing temporary files that can be used to populate configuration templatesF 1 Including nodesN 5 Node number, node index, clusterC 6 Cluster number, cluster index, cluster name; filling node types by Golang template filling technique using temporary filesT 3 All configuration templates under the node are obtained to be suitable for the nodeN 5 Is a configuration file of (a).
2. A distributed system configuration generation method for a securities trading system according to claim 1, wherein the specific sequence of steps is:
step S1, constructing a source file for generating a configuration file, wherein the source file comprises an infrastructure information file, deployment information, a service statement file and a configuration file template;
s2, using a configuration center client to upload all configuration schemes and package a source file to a configuration center server;
s3, immediately performing one-time generation test on the source file after the server receives the source file;
step S4, if the verification in the step 3 is correct, storing the uploaded source file packet into a high-availability database, distributing a globally unique version number for the source file packet, returning the version number to a request client, recording the version number into the high-availability database, and if the verification is not correct, returning an error;
step S5, when the client requests to generate the configuration file, firstly querying the version number of the source file package existing in the server, selecting a target version number, and sending a generation request to the server;
and S6, the server generates a configuration file package according to the step 3 and returns the configuration file package to the client.
3. The method for generating distributed system configuration suitable for securities trading system according to claim 1, wherein in the step S1, an infrastructure information file is constructed, specifically, information of each available physical machine is collected, including a host name, all network names, network card names located in different networks, IPv4 addresses and IPv6 addresses bound by each network card, and the position of the host is written as a file by using any text serialization format; the construction and deployment information specifically comprises the following steps: constructing a deployment information file by a cluster level, wherein the number of nodes of the cluster and the name of a host computer to be deployed by each node are required to be indicated in the file; the construction of the service statement file is specifically as follows: constructing a service declaration file at the node type level for indicating the type of communication information, the file comprising three parts, one being intra-cluster channels, the part having to indicate on which network the node type communicates; secondly, the service channel, the part must specify the service channel to be published and the name of the service channel to be subscribed to; thirdly, a proprietary channel, wherein the part must specify the node type of a receiver of the proprietary channel; the construction of the configuration file template specifically comprises the following steps: and constructing a configuration file template at the node type level, generating actual configuration files of all nodes under the node type by virtue of Golang template filling technology, wherein available replacement items on the configuration file template comprise cluster names, cluster numbers, cluster indexes, node numbers and node indexes, and filling positions requiring replacement on the template by using actual deployment information of the node when generating the configuration file of a certain node.
4. A distributed system configuration generation method suitable for a securities trading system according to claim 3, wherein multiple sets of conventional configuration schemes can be prepared at the same time according to actual needs, and step S1 is repeated until the setting of all the configuration schemes is completed, and the schemes are placed in the same configuration file package.
5. The method of claim 1, wherein the step S31 further comprises creating a deployment information table, clusterC 1 After the unprocessed node is selected, the corresponding host information is found on the infrastructure information file according to the host name, the node number, the node index and the host information of the node are recorded in the deployment information table, and if the host name does not appear on the infrastructure information file, the error is reported back.
6. The method of claim 1, wherein said step (1) is performed by cycling through all nodes in all clusters in the plan P.
7. A distributed system configuration generation method for use in a securities trading system according to claim 1, wherein said step (2) loops through until node type T 1 All the clusters in the network complete the generation of multicast channels and node types in the clustersT 1 All clusters in (1) have been assigned a claimB 1 Corresponding multicast channel and node typeT 1 All network names in (1)W 2 The following business statement is distributed with multicast channel and clusterC 4 All nodes in the network have recorded a statementB 2 Multicast information of (3) clusterC 5 All nodes in the network have recorded a statementB 2 Multicast information up to sender clusterC 4 Corresponding proprietary declarations with all recipient clustersB 2 Is distributed and recorded until the node typeT 1 All clusters in (1) correspond to declarationsB 2 Is distributed and recorded until the node typeT 1 All network names in (1)W 2 Proprietary declarations under the network have all been assigned multicast channels up to the node typeT 1 The service type declaration and the proprietary declaration under all networks are distributed and recorded completely until the communication information of all node types in the scheme P is processed and recorded completely.
8. A distributed system configuration generation method for use in a securities trading system according to claim 1, wherein said step (3) loops through to clustersC 6 All nodes in the network generate corresponding configuration files until the node type T 3 Corresponding configuration files are generated by all nodes in the scheme P until the corresponding configuration files are generated by all nodes in the scheme P, until all configuration schemes in the source file package pass verification, and if errors occur, the errors are returned.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210776976.2A CN115208883B (en) | 2022-07-04 | 2022-07-04 | Distributed system configuration generation method suitable for securities trading system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210776976.2A CN115208883B (en) | 2022-07-04 | 2022-07-04 | Distributed system configuration generation method suitable for securities trading system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115208883A CN115208883A (en) | 2022-10-18 |
| CN115208883B true CN115208883B (en) | 2023-09-12 |
Family
ID=83577609
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210776976.2A Active CN115208883B (en) | 2022-07-04 | 2022-07-04 | Distributed system configuration generation method suitable for securities trading system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115208883B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2000062502A2 (en) * | 1999-04-12 | 2000-10-19 | Rainfinity, Inc. | Distributed server cluster for controlling network traffic |
| US6718361B1 (en) * | 2000-04-07 | 2004-04-06 | Network Appliance Inc. | Method and apparatus for reliable and scalable distribution of data files in distributed networks |
| CN114185577A (en) * | 2021-12-16 | 2022-03-15 | 广州市百果园网络科技有限公司 | Configuration management system, configuration management method, server, and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6950833B2 (en) * | 2001-06-05 | 2005-09-27 | Silicon Graphics, Inc. | Clustered filesystem |
-
2022
- 2022-07-04 CN CN202210776976.2A patent/CN115208883B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2000062502A2 (en) * | 1999-04-12 | 2000-10-19 | Rainfinity, Inc. | Distributed server cluster for controlling network traffic |
| US6718361B1 (en) * | 2000-04-07 | 2004-04-06 | Network Appliance Inc. | Method and apparatus for reliable and scalable distribution of data files in distributed networks |
| CN114185577A (en) * | 2021-12-16 | 2022-03-15 | 广州市百果园网络科技有限公司 | Configuration management system, configuration management method, server, and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 基于Zookeeper的配置管理中心设计与实现;苗凡;阎志远;戴琳琳;;铁路计算机应用(10);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115208883A (en) | 2022-10-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111240763B (en) | Configuration updating method, device, equipment and storage medium | |
| CN107395353B (en) | block chain consensus method and device | |
| CN101504672B (en) | Method and system for dynamically configuring physical data table | |
| CN108600148B (en) | Transaction message processing method and device | |
| CN101908015B (en) | Device and method for creating test case based on components | |
| CN101661584B (en) | Dynamic order workflow template instantiator and decoupler | |
| CN108509392B (en) | Multi-mechanism account checking method, system, computer equipment and storage medium | |
| UA65638C2 (en) | Method and system for synchronization and management of a data base | |
| CN104572122A (en) | Software application data generating device and method | |
| KR102172514B1 (en) | Managing method for test data based on blockchain node apparatus of blockchain | |
| CN105630590A (en) | Business information processing method and device | |
| US20230289782A1 (en) | Smart contract-based data processing | |
| NZ531849A (en) | Information object authoring and distribution system | |
| WO2021195618A1 (en) | System and method for integration and validation | |
| CN111949726A (en) | Relational database synchronization method and system based on block chain | |
| CN106850724A (en) | Data push method and device | |
| WO2009147705A1 (en) | Database connection program and device | |
| CN115208883B (en) | Distributed system configuration generation method suitable for securities trading system | |
| CN112363997B (en) | Data version management method, device and storage medium | |
| CN108376082A (en) | A kind of reference library method for updating edition and system | |
| CN113220633A (en) | Unified file coding management method and system | |
| CN107341645A (en) | Demand of human resources information processing method, device, computer equipment and storage medium | |
| CN106656592A (en) | Service management method and device based on role configuration | |
| CN114598680B (en) | Domain name management method, device and storage medium | |
| CN106506661A (en) | Method, server and system for dynamically returning data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |