CN112613860B

CN112613860B - Deep traffic institute TGW-based cluster disk reporting system

Info

Publication number: CN112613860B
Application number: CN202011500389.8A
Authority: CN
Inventors: 曹世荣; 陈利; 宋屠康; 单兴邦; 王刚
Original assignee: Orient Securities Co ltd
Current assignee: Orient Securities Co ltd
Priority date: 2020-12-17
Filing date: 2020-12-17
Publication date: 2022-08-09
Anticipated expiration: 2040-12-17
Also published as: CN112613860A

Abstract

The invention relates to a cluster report system based on a deep crossing bureau TGW, which comprises a high-speed transaction bus and a report service module, wherein the high-speed transaction bus is communicated with a transaction engine through a UDP multicast network, receives a transaction order of the transaction engine and sends the transaction order to the report service module, and sends confirmation and successful delivery report data to the transaction engine at the same time, the report service module comprises a bus processing sub-module, a business processing sub-module and a plurality of TCP connections, the bus processing sub-module receives the transaction order, compiles the transaction orders belonging to the same report group into the same order queue, and the business processing sub-module reports the order queue to the deep crossing bureau TGW through the TCP connections according to a round-robin order report algorithm. Compared with the prior art, the invention has the advantages of improving the utilization rate of the report seat, effectively solving the problem of overstock order when entrusting report, and the like.

Description

Deep traffic institute TGW-based cluster disk reporting system

Technical Field

The invention relates to the field of financial science and technology, in particular to a cluster report system based on a deep traffic institute TGW.

Background

The hosting mechanism of the deep traffic place is automatic hosting, and buying and selling are carried out everywhere without limitation. However, the transaction has a very high requirement on real-time performance, the multi-seat transaction causes complex ticket verification and ticket checking and freezing and unfreezing algorithms, the transaction performance is greatly reduced, meanwhile, the multi-seat transaction causes difficulty in clearing, and often, a customer only has one hosting seat in one security dealer and is only connected with one hosting Transaction Gateway (TGW) for quotation during transaction. Seat resources are important channel resources of security dealer, and the method that a plurality of clients use the same seat is the most common method of security dealer, and especially a large number of scattered households share one seat. But a direct problem with seat sharing is the sharing of traffic. Flow rate control is very disadvantageous if the customer is doing high frequency transactions. But subject to the flow weight limit of a single TGW, even if a single client shares one TGW alone, is limited to the TGW maximum flow weight. Expanding flow rates are a necessary appeal to high frequency trading customers.

In the existing single seat and single TGW mode, if the TGW fails, the standby TGW mode needs to be started urgently, the report service is interrupted briefly, and if the server where the TGW is located fails seriously, the main-standby switching usually needs longer time. Highly available designs are an important concept for dealer core system design, but many highly available designs are often difficult to implement due to difficult engineering implementation, or are difficult to scale up due to prohibitive cost. Therefore, the online fault isolation method is especially necessary for the situation that the TGW fails or the machine where the TGW is located fails to work.

If a single client shares one seat and shares one group of TGWs for reporting, when the client is in low peak, a plurality of TGWs will cause huge waste. If a plurality of customers with transaction peaks in different time periods share a group of TGWs for transaction, the resources of the TGWs can be greatly saved when the TGWs are rare even if the customers are all located on the same seat.

Disclosure of Invention

The invention aims to overcome the defects of the prior art and provide a cluster report system based on a deep traffic bureau TGW, which meets the requirement of flow rate in high-frequency transaction, effectively solves the problem of resource waste caused by uneven flow rate distribution of a plurality of clients in time, and ensures high-availability and low-delay top-speed report.

The purpose of the invention can be realized by the following technical scheme:

a cluster report system based on a deep transit station TGW comprises a high-speed transaction bus and a report service module, wherein the high-speed transaction bus is communicated with a transaction engine through a UDP multicast network, receives a transaction order of the transaction engine and sends the transaction order to the report service module, and meanwhile sends confirmation and completion report data to the transaction engine, the report service module comprises a bus processing sub-module, a business processing sub-module and a plurality of TCP connections, the bus processing sub-module receives the transaction order and compiles the transaction orders belonging to the same report group into the same order queue, and the business processing sub-module reports the order queue to the deep transit station TGW through the TCP connections according to an order report algorithm.

The bus processing submodule comprises a data analysis component and a queuing processing component.

Furthermore, the data analysis component analyzes the transaction data received by the bus processing sub-module into a transaction order through a protobuf data analysis technology, and the queuing processing component compiles the transaction order into an order queue according to a time sequence.

Further, the order queue is specifically a linear table structure of transaction order data.

Further, the data analysis component analyzes the bottom binary data and packages the data according to a preset order data structure to form the transaction order.

The business processing submodule is provided with a plurality of transaction processing units, the number of the transaction processing units is the same as that of the transaction units corresponding to the quotation group, the business processing submodule is classified one by one according to the transaction units, and the transaction orders corresponding to the same transaction unit are delivered to the corresponding transaction processing units.

Further, the transaction processing unit receives the order queue sent by the transaction unit of the corresponding report group from the service processing submodule, splits the order queue according to the round-robin and the preset unit sequence, and circularly reports the split order queue to the deep traffic bureau TGW through TCP connection.

Furthermore, a corresponding TGW connection state list is arranged in the cluster reporting system, all TGW healthy connection states are monitored in a heartbeat monitoring mode, if one TGW in the list is abnormal, the monitored abnormal TGW is removed from the TGW healthy state list, and the connection with the abnormal TGW is disconnected. And the received order queue is re-split according to the healthy TGW list, and the split order queue is circularly reported to the deep transit TGW through TCP connection.

Furthermore, the function of the cluster reporting system further includes receiving order confirmation data and transaction return data returned by the deep drafts TGW, identifying the order confirmation data according to the reporting contract number of the transaction order, filtering the repeatedly received order confirmation data, identifying the transaction return data according to the reporting contract number of the transaction order and the transaction number of the deep drafts, and filtering the repeatedly received transaction return data, so as to prevent the transaction engine from repeatedly processing order confirmation and transaction return, and avoid the repeated calculation of funds and securities.

The round-robin distribution algorithm is provided with a round-robin mechanism and a fault identification and isolation mechanism.

The high-speed transaction bus is based on a high-speed ten-trillion network card and is in the same ten-trillion multicast network with the transaction engine through a high-speed switch.

Compared with the prior art, the invention has the following beneficial effects:

1. the invention sets the high-speed transaction bus, and the high-speed transaction bus and the transaction engine are in the same trillion multicast network, thereby meeting the technical requirements of high availability, low delay and the like of the system and providing a technical basis for the rapid report.

2. According to the invention, through the cluster type report service module, each TGW is bound with one seat, and N TGWs form a report disc group, so that report disc alternation of multiple seats is realized, the problem of resource waste caused by uneven distribution of flow velocity at different time is effectively solved, the utilization rate of report seats is greatly improved, the expansion of transaction units can be realized, the effective utilization of resources is realized, and the cost is saved.

3. The order queue is split through the sequential reporting service, the order sequence of a single queue is respectively corresponding to the multiple TGWs for reporting according to the sequence, the highest flow rate right of reporting flow rate is expanded, the problem of order overstock during entrusting reporting is effectively solved, the reporting rate is greatly improved, and the trading requirement of fast forward and fast forward of investors under the instantaneous peak value is met.

4. The method adopts a heartbeat monitoring mode to monitor one TGW in all the TGW health state lists, such as abnormal connection, eliminates the monitored abnormal TGW from the TGW health state lists, and re-splits an order queue according to the healthy TGW lists after the abnormal TGW is eliminated, so that the failure isolation without perception of a user system is realized.

Drawings

FIG. 1 is a schematic view of the structure of the present invention;

FIG. 2 is a schematic diagram of a sequential reporting in an embodiment of the present invention;

fig. 3 is a schematic flow chart of the sequential reporting according to the embodiment of the present invention.

Detailed Description

The invention is described in detail below with reference to the figures and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.

Examples

As shown in fig. 1, a cluster reporting system based on deep hand exchange TGW includes a high-speed transaction bus and a reporting service module, where the high-speed transaction bus communicates with a transaction engine through a UDP multicast network, receives a transaction order from the transaction engine, sends the transaction order to the reporting service module, and sends confirmation and transaction return data to the transaction engine, and the reporting service module includes a bus processing sub-module, a service processing sub-module, and a plurality of TCPs connected to each other, where the bus processing sub-module receives the transaction order, and compiles the transaction orders belonging to the same reporting group into the same order queue, and the service processing sub-module reports the order queue to the deep hand exchange TGW through the TCPs according to a round-robin algorithm.

The data analysis component analyzes the transaction data received by the bus processing sub-module into a transaction order through a protobuf data analysis technology, and the queuing processing component compiles the transaction order into an order queue according to the time sequence.

The order queue is specifically a linear table structure of trade order data.

And the data analysis component analyzes the bottom layer binary data of the transaction data and packages the data according to a preset order data structure to form a transaction order.

The business processing submodule is provided with a plurality of transaction processing units, the number of the transaction processing units is the same as that of the transaction units corresponding to the quotation group, the business processing submodule is classified one by one according to the transaction units, and the transaction orders corresponding to the same transaction unit are delivered to the corresponding transaction processing unit.

And the transaction processing unit receives the order queue sent by the transaction unit of the corresponding report disk group from the service processing submodule, splits the order queue according to the round-robin report algorithm and the preset unit sequence, and circularly reports the split order queue to the deep traffic bureau TGW through the TCP.

The cluster reporting system is provided with a corresponding TGW health state list, monitors the TGW health connection state in a heartbeat monitoring mode, such as one TGW in the connection abnormal list, eliminates the monitored abnormal TGW from the TGW health state list, and disconnects the connection with the fault TGW. And after the order queue is removed, the received order queue is re-split according to the healthy TGW list, and the split order queue is circularly reported to the deep traffic station TGW through TCP connection.

The cluster reporting system also comprises the functions of receiving order confirmation data and transaction return data returned by the deep floor TGW, identifying the order confirmation data according to the reporting contract number of the transaction order, filtering the repeatedly received order confirmation data, identifying the transaction return data according to the reporting contract number of the transaction order and the transaction number of the deep floor, and filtering the repeatedly received transaction return data, so that a transaction engine is prevented from repeatedly processing order confirmation and transaction return, and the repeated calculation of funds and securities is avoided.

The round-sequence disk-reporting algorithm is provided with a round-sequence mechanism and a fault identification and isolation mechanism.

As shown in fig. 2, where X is a hosting seat, a1, a2, and A3 are reporting seats, and the round-robin scheduling algorithm is based on the deep-dealers seat and TGW decoupling (different TGWs bind to the same seat), so as to ensure that the order of the same transaction unit can be reported to the deep-dealers through different TGWs, and the orders of the transactions in the order sequence of different transaction units in the scheduling service are sequentially reported among multiple TGWs in a time sequence, thereby ensuring uniform distribution of flow rate, improving resource utilization rate, and avoiding resource waste.

As shown in fig. 3, in this embodiment, the number of TGW resources is 3, and the method specifically includes the following steps when performing round-robin distribution of trade orders:

step S1: the system transaction engine receives transaction orders reported by various terminals, namely an order 1, an order 2, an order 3, an order 4 and an order 5;

step S2: the trading engine transmits order data of the order 1, the order 2, the order 3, the order 4 and the order 5 to a high-speed trading bus in sequence;

step S3: the high-speed trading bus transmits order data of order data, namely an order 1, an order 2, an order 3, an order 4 and an order 5 to the report service module, a bus processing submodule arranged in the report service module performs binary order data analysis to respectively generate an order in a structural data form (I), an order in a structural data form (II), an order in a structural data form (III), an order in a structural data form (IV) and an order in a structural data form (V), and the order data are respectively put into order sequences of corresponding trading units according to the trading units;

step S4: the business processing sub-module built in the report service module sequentially sends orders in the order sequence to corresponding transaction processing units according to different transaction units, and then carries out sequential report, namely, the orders are reported to TGW1, the orders are reported to TGW2, the orders are reported to TGW3, the orders are reported to TGW1, the orders are reported to TGW2, and the orders are sequentially reported in sequence according to the sequence;

step S5: after the deep delivery receives the order, after the deep delivery finishes processing the order, the confirmation and return information of the order are fed back to the TGW from the TGW which is the same as the report, and the confirmation and return information is reversely returned to the transaction engine through the original report path to finish the report.

In step S4, if the order is reported to TGW2, TGW2 fails, and the order is reported to TGW3 in turn.

In addition, it should be noted that the specific implementation examples described in this specification may have different names, and the above contents described in this specification are only illustrations of the structures of the present invention. All equivalent or simple changes in the structure, characteristics and principles of the invention are included in the protection scope of the invention. Various modifications or additions may be made to the described embodiments or methods may be similarly employed by those skilled in the art without departing from the scope of the invention as defined in the appending claims.

Claims

1. A cluster report system based on a deep transaction station TGW is characterized by comprising a high-speed transaction bus and a report service module, wherein the high-speed transaction bus is communicated with a transaction engine through a UDP multicast network, receives a transaction order of the transaction engine, sends the transaction order to the report service module, and simultaneously sends confirmation and transaction report data to the transaction engine;

the business processing submodule is provided with a plurality of transaction processing units, the number of the transaction processing units is the same as that of the transaction units corresponding to the quotation group, and the business processing submodule is used for classifying the transaction units one by one and delivering the transaction orders corresponding to the same transaction unit to the corresponding transaction processing unit;

and the transaction processing unit receives the order queue sent by the transaction unit of the corresponding report disk group from the service processing submodule, splits the order queue according to the round-robin and the preset unit sequence, and circularly reports the split order queue to different TGWs of the deep traffic station through TCP connection.

2. The deep traffic bureau TGW-based cluster quotation system of claim 1, wherein the bus processing submodule comprises a data parsing component and a queuing processing component.

3. The deep traffic bureau TGW-based cluster quotation system of claim 2, wherein the data parsing component parses the transaction data received by the bus processing sub-module into transaction orders, and the queuing processing component queues the transaction orders in an order queue according to a chronological order.

4. The deep traffic bureau TGW-based cluster quotation system of claim 3, wherein the data parsing component is configured to form the trade order by parsing underlying binary data and packaging the binary data according to a predetermined order data structure.

5. The TGW-based cluster reporting system of claim 1, wherein the cluster reporting system is provided with a corresponding TGW connection status list, monitors the health status of each TGW in the TGW connection status list by means of heartbeat monitoring, and removes the abnormal TGW from the TGW health list to disconnect the abnormal TGW.

6. The TGW-based clustered listing system of claim 5, wherein the functions of the TGW-based clustered listing system further comprise receiving order confirmation data and return-to-transaction data returned from the TGW, identifying the order confirmation data according to the listing contract number of the transaction order, filtering the repeatedly received order confirmation data, identifying the return-to-transaction data according to the listing contract number of the transaction order and the return-to-transaction number of the deep transaction, and filtering the repeatedly received return-to-transaction data.

7. The deep traffic bureau TGW-based cluster reporting system according to claim 1, wherein the round-robin ordering algorithm is provided with a round-robin ordering mechanism and a fault identification and isolation mechanism.

8. The deep traffic bureau TGW-based cluster quotation system of claim 1, wherein the high-speed transaction bus is based on a high-speed ten-trillion network card, and is in the same ten-trillion multicast network with the transaction engine through a high-speed switch.