CN106934031B

CN106934031B - Method and device for monitoring and processing hotspot records in real-time processing system

Info

Publication number: CN106934031B
Application number: CN201710150032.3A
Authority: CN
Inventors: 商建祥; 王继武; 张国华
Original assignee: Bank of China Ltd
Current assignee: Bank of China Ltd
Priority date: 2017-03-14
Filing date: 2017-03-14
Publication date: 2020-03-13
Anticipated expiration: 2037-03-14
Also published as: CN106934031A

Abstract

The invention provides a method and a device for monitoring and processing hotspot records in a real-time processing system, wherein the method comprises the following steps: calculating to obtain response time according to the receiving time stamp of the transaction request packet and the return time stamp of the transaction request packet after being processed by the real-time processing system; judging whether the response time exceeds a set threshold value processing time or not; if yes, updating a performance data table according to the record information in the request packet and the return timestamp, wherein the performance data table comprises a hotspot record information column, a statistical window ending time column and an abnormal transaction number column in a statistical window; judging whether a hot spot appears on the transaction type corresponding to the request packet and the record information according to the data of the abnormal transaction number column in the statistical window in the updated performance data table and a set hot spot threshold value; if so, limiting the transaction for the transaction type and for the logging information. The invention can accurately position and process the hotspot record.

Description

Method and device for monitoring and processing hotspot records in real-time processing system

Technical Field

The invention relates to the technical field of computers, in particular to a method and a device for monitoring and processing hotspot records in a real-time processing system.

Background

Real-time processing systems typically allow for multi-process concurrent or multi-thread concurrent processing, and thus have powerful transaction concurrent processing capabilities, and typically can further increase the concurrent processing capabilities by expanding the number of processes/threads. In a real-time processing system, many processes/threads work in parallel, but each uses its own resources, each processing its own data, as if the car were running on a multi-lane highway. However, when a plurality of processes/threads need to process the same record, a situation of waiting for each other occurs, that is, a record hot spot occurs, and as lanes of a multi-lane highway are suddenly reduced, a multi-lane vehicle needs to rob one lane, and at this time, the processing speed of the related process/thread is reduced.

The occurrence of a recording hot spot may quickly lead to a decrease in processing performance of the related transaction (the transaction that needs to operate the recording hot spot), i.e., a congestion phenomenon. Furthermore, these transactions may lock other resources or records, such as waiting transactions to fully fill the threads of the real-time processing system, transactions that cannot be processed even without accessing hot-spot records, and congestion may propagate until the real-time processing system is down.

The primary problem of solving the recording hotspot is to locate. The general locating means is to analyze the use condition and time consumption of each table and index through a database tool to confirm the specific hotspot record. However, this positioning means is performed afterwards, and only provides a basis for post-program optimization, such as reducing accesses to one record collectively for a plurality of times from processing logic. The solution to the problem of system congestion caused by hot spot logging is to limit the current. The common current limiting means is to limit the current of the transaction request of the whole system or limit the current of some kind of transaction. However, these current-limiting means are not only ineffective, but also are all bystander and innocent due to inaccurate current-limiting.

Disclosure of Invention

The invention provides a method and a device for monitoring and processing hotspot records in a real-time processing system, which are used for accurately positioning and disposing the hotspot records.

The invention provides a method for monitoring and processing hotspot records in a real-time processing system, which comprises the following steps: receiving a transaction request packet and recording a receiving time stamp of the transaction request packet; receiving a transaction return packet returned after the real-time processing system processes the transaction request packet, and recording a return timestamp of the transaction return packet; calculating the transaction response time according to the receiving timestamp and the returning timestamp; judging whether the transaction response time exceeds a set threshold processing time or not; if so, updating a performance data table according to the record information in the transaction request packet and the return timestamp, wherein the performance data table comprises a hotspot record information column corresponding to the record information, a statistic window ending time column and an abnormal transaction number column in a statistic window; judging whether a hot spot appears on the transaction type corresponding to the transaction request packet and the record information according to the updated data of the abnormal transaction number column in the statistical window in the performance data table and a set hot spot threshold value; and if so, limiting the transaction related to the transaction type and the record information according to set current limiting parameters.

In one embodiment, before receiving a transaction return packet returned after the real-time processing system processes the transaction request packet, the method further includes: judging whether the flow limitation needs to be carried out on the transaction corresponding to the transaction request packet or not according to the data of the abnormal transaction number column in the statistical window in the performance data table before updating and the set hot spot threshold value; if not, the transaction request packet is forwarded to the real-time processing system for logic processing.

In one embodiment, before determining whether to perform current limiting on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot spot threshold, the method further includes: judging whether a performance data record related to the record information exists in the performance data by searching the performance data table; if not, the step of forwarding the transaction request packet to the real-time processing system for logic processing is executed; if yes, judging whether the performance data record of the record information meets the following conditions: and if not, forwarding the transaction request packet to the real-time processing system for logic processing, and if so, judging whether the flow limitation needs to be performed on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot point threshold value.

In one embodiment, the method further comprises: if the transaction corresponding to the transaction request packet needs to be limited, judging whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in a current limiting window corresponding to the record information in the performance data table before updating; if yes, generating a transaction rejection packet according to the transaction request packet, taking the time point generated by the transaction rejection packet as a return timestamp, and executing the step of updating a performance data table according to the record information in the transaction request packet and the return timestamp; if not, updating the data of the forwarded transaction number column in the current limiting window corresponding to the record information in the performance data table, and executing the step of forwarding the transaction request packet to the real-time processing system for logic processing.

In one embodiment, before determining whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in the current limiting window corresponding to the record information in the performance data table before updating, the method further includes: and judging whether the current time point is less than or equal to the data of the current limiting window ending time column corresponding to the record information in the performance data table, if not, updating the data of the current limiting window ending time column according to the current time point and the set current limiting window duration.

In one embodiment, the performance data table includes a plurality of abnormal transaction number fields in the statistical window, and the statistical windows of the abnormal transaction number fields in the statistical windows are arranged in time sequence.

In one embodiment, updating the performance data table according to the record information in the transaction request packet and the return timestamp includes: acquiring a statistical window ending time column and an abnormal transaction quantity column in a statistical window related to the record information in the performance data table; calculating new statistical window ending time according to the return timestamp and the set statistical window duration so that the return timestamp is smaller than the new statistical window ending time, and updating the data of the acquired statistical window ending time column by using the new statistical window ending time; calculating the new end time of the statistical window of the abnormal transaction quantity column in each acquired statistical window according to the new end time of the statistical window and the set statistical window duration; comparing whether the new end time is the same as the window end time corresponding to the abnormal transaction number column in the obtained statistical window or not; if the two end times are the same, updating the data of the abnormal transaction number field in the obtained statistical window corresponding to the window end time by using the abnormal transaction number field in the obtained statistical window corresponding to the new end time; and adding one to the data of the abnormal transaction number field in the statistical window acquired at the last time, and updating the data of the abnormal transaction number field in the acquired statistical window which is not updated to zero.

In one embodiment, the obtaining the field of the end time of the statistic window and the field of the number of abnormal transactions in the statistic window related to the record information in the performance data table includes: judging whether the performance data table contains a hotspot recording information column corresponding to the recording information or not by searching the performance data table; if yes, obtaining a statistical window ending time field and an abnormal transaction quantity field in the statistical window related to the record information according to the searched hotspot record information field.

In one embodiment, the obtaining a field of a statistical window ending time and a field of a number of abnormal transactions in a statistical window related to the logging information in the performance data table further includes: if the performance data table does not contain the hotspot record information column corresponding to the record information, adding the hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window related to the record information in the performance data table, and initializing the added hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window.

In one embodiment, the obtaining a field of a statistical window ending time and a field of a number of abnormal transactions in a statistical window related to the logging information in the performance data table further includes: if the performance data table does not contain the hotspot record information column corresponding to the record information, searching and finding a statistic window failure record in the performance data table, initializing the statistic window failure record, and respectively taking the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the initialized statistic window failure record as the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the record information, wherein the statistic window failure record satisfies the following conditions: and returning a timestamp > the end time of the statistical window in the statistical window failure record + the number of the set statistical windows and the set statistical window duration.

In one embodiment, the determining whether a hot spot occurs in the transaction type and the record information corresponding to the transaction request packet according to the updated data of the abnormal transaction number field in the statistical window in the performance data table and the set hot spot threshold includes: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the updated performance data table exceeds the set hot spot threshold value, and if so, confirming that a hot spot appears on the transaction type and the record information.

In one embodiment, the performance data table includes a plurality of abnormal transaction number fields in the statistical window, and the statistical windows of the abnormal transaction number fields in each statistical window are arranged in time sequence; judging whether the flow limitation needs to be performed on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot spot threshold value, wherein the judging step comprises the following steps: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the performance data table before updating exceeds the set hot spot threshold value, if so, determining that the current limitation needs to be performed on the transaction corresponding to the transaction request packet.

In one embodiment, the setting the current limiting parameter includes: the time length of the current-limiting window, the number of the current-limiting windows and the number of transactions allowed to be forwarded by each current-limiting window; the performance data table further comprises a forwarded transaction number field in a plurality of current limiting windows, the number of the forwarded transaction number fields being the number of the current limiting windows, and the current limiting windows of the forwarded transaction number fields in each current limiting window have the duration of the current limiting window and are arranged in time sequence; judging whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in the current limiting window corresponding to the record information in the performance data table before updating, wherein the judging comprises the following steps: and judging whether the real-time processing system needs to be prevented from processing the transaction request packet or not according to whether the data of the forwarded transaction number field in each current limiting window corresponding to the record information in the performance data table before updating is larger than or equal to the number of the transactions allowed to be forwarded by the current limiting window.

In one embodiment, the transaction type is represented by data of a transaction type encoding field, and the transaction type encoding field is pre-added to the transaction request packet.

In one embodiment, the record information is obtained according to an input packet start bit and an input packet end bit.

The invention also provides a device for monitoring and processing hotspot records in a real-time processing system, which comprises: a request packet interception unit for: receiving a transaction request packet and recording a receiving time stamp of the transaction request packet; a return packet interception unit for: receiving a transaction return packet returned after the real-time processing system processes the transaction request packet, and recording a return timestamp of the transaction return packet; a response time calculation unit for: calculating the transaction response time according to the receiving timestamp and the returning timestamp; a collected information judging unit configured to: judging whether the transaction response time exceeds a set threshold processing time or not; a performance data update unit to: if so, updating a performance data table according to the record information in the transaction request packet and the return timestamp, wherein the performance data table comprises a hotspot record information column corresponding to the record information, a statistic window ending time column and an abnormal transaction number column in a statistic window; a hotspot record judging unit, configured to: judging whether a hot spot appears on the transaction type corresponding to the transaction request packet and the record information according to the updated data of the abnormal transaction number column in the statistical window in the performance data table and a set hot spot threshold value; a hotspot record processing unit, configured to: and if so, limiting the transaction related to the transaction type and the record information according to set current limiting parameters.

In one embodiment, the method further comprises: a transaction flow limit determination unit configured to: judging whether the flow limitation needs to be carried out on the transaction corresponding to the transaction request packet or not according to the data of the abnormal transaction number column in the statistical window in the performance data table before updating and the set hot spot threshold value; a request packet forwarding unit configured to: if not, the transaction request packet is forwarded to the real-time processing system for logic processing.

In one embodiment, the method further comprises: a performance data search unit to: judging whether a performance data record related to the record information exists in the performance data by searching the performance data table; a performance data validity judgment unit configured to: if not, the transaction request packet is forwarded to the real-time processing system for logic processing; if yes, judging whether the performance data record of the record information meets the following conditions: and if not, forwarding the transaction request packet to the real-time processing system for logic processing, and if so, judging whether the flow limitation needs to be performed on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot point threshold value.

In one embodiment, the method further comprises: a transaction block determination unit to: if the transaction corresponding to the transaction request packet needs to be limited, judging whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in a current limiting window corresponding to the record information in the performance data table before updating; a transaction blocking processing unit to: if yes, generating a transaction rejection packet according to the transaction request packet, taking the time point generated by the transaction rejection packet as a return timestamp, and updating a performance data table according to the record information in the transaction request packet and the return timestamp; if not, updating the data of the forwarded transaction number column in the current limiting window corresponding to the record information in the performance data table, and executing to forward the transaction request packet to the real-time processing system for logic processing.

In one embodiment, the method further comprises: a current limit window end time update unit for: and judging whether the current time point is less than or equal to the data of the current limiting window ending time column corresponding to the record information in the performance data table, if not, updating the data of the current limiting window ending time column according to the current time point and the set current limiting window duration.

In one embodiment, the performance data updating unit is further configured to: the performance data table comprises a plurality of abnormal transaction quantity columns in the statistical window, and the statistical windows of the abnormal transaction quantity columns in the statistical windows are arranged in a time sequence.

In one embodiment, the performance data updating unit includes: a registration-related field acquisition module configured to: acquiring a statistical window ending time column and an abnormal transaction quantity column in a statistical window related to the record information in the performance data table; a first field data update module configured to: calculating new statistical window ending time according to the return timestamp and the set statistical window duration so that the return timestamp is smaller than the new statistical window ending time, and updating the data of the acquired statistical window ending time column by using the new statistical window ending time; a second field data update module configured to: calculating the new end time of the statistical window of the abnormal transaction quantity column in each acquired statistical window according to the new end time of the statistical window and the set statistical window duration; comparing whether the new end time is the same as the window end time corresponding to the abnormal transaction number column in the obtained statistical window or not; if the two end times are the same, updating the data of the abnormal transaction number field in the obtained statistical window corresponding to the window end time by using the abnormal transaction number field in the obtained statistical window corresponding to the new end time; and adding one to the data of the abnormal transaction number field in the statistical window acquired at the last time, and updating the data of the abnormal transaction number field in the acquired statistical window which is not updated to zero.

In one embodiment, the registration-related field obtaining module includes: the hotspot record column searching module is used for: judging whether the performance data table contains a hotspot recording information column corresponding to the recording information or not by searching the performance data table; a registration-related field acquisition module configured to: if yes, obtaining a statistical window ending time field and an abnormal transaction quantity field in the statistical window related to the record information according to the searched hotspot record information field.

In one embodiment, the record-related field obtaining module further includes: a record-related field addition module configured to: if the performance data table does not contain the hotspot record information column corresponding to the record information, adding the hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window related to the record information in the performance data table, and initializing the added hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window.

In one embodiment, the record-related field obtaining module further includes: a failure record recovery module to: if the performance data table does not contain the hotspot record information column corresponding to the record information, searching and finding a statistic window failure record in the performance data table, initializing the statistic window failure record, and respectively taking the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the initialized statistic window failure record as the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the record information, wherein the statistic window failure record satisfies the following conditions: and returning a timestamp > the end time of the statistical window in the statistical window failure record + the number of the set statistical windows and the set statistical window duration.

In one embodiment, the hot spot record determining unit includes: the hotspot record judging module is used for: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the updated performance data table exceeds the set hot spot threshold value, and if so, confirming that a hot spot appears on the transaction type and the record information.

In one embodiment, the performance data table includes a plurality of abnormal transaction number fields in the statistical window, and the statistical windows of the abnormal transaction number fields in each statistical window are arranged in time sequence; the transaction flow limit judging unit comprises: a transaction flow limit determination module to: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the performance data table before updating exceeds the set hot spot threshold value, if so, determining that the current limitation needs to be performed on the transaction corresponding to the transaction request packet.

In one embodiment, the setting the current limiting parameter includes: the time length of the current-limiting window, the number of the current-limiting windows and the number of transactions allowed to be forwarded by each current-limiting window; the performance data table further comprises a forwarded transaction number field in a plurality of current limiting windows, the number of the forwarded transaction number fields being the number of the current limiting windows, and the current limiting windows of the forwarded transaction number fields in each current limiting window have the duration of the current limiting window and are arranged in time sequence; the transaction block judgment unit includes: a transaction block determination module to: and judging whether the real-time processing system needs to be prevented from processing the transaction request packet or not according to whether the data of the forwarded transaction number field in each current limiting window corresponding to the record information in the performance data table before updating is larger than or equal to the number of the transactions allowed to be forwarded by the current limiting window.

In one embodiment, the performance data updating unit is further configured to: the transaction type is represented by data of a transaction type encoding field, which is pre-added to the transaction request packet.

In one embodiment, the performance data updating unit is further configured to: and the recording information is obtained according to the start bit and the end bit of the input packet.

According to the monitoring and processing method and device for the hot spot record in the real-time processing system, the performance data table is updated according to the record information and the return timestamp in the transaction request packet, whether the hot spot occurs on the transaction type and the record information corresponding to the transaction request packet is judged according to the data of the abnormal transaction number field in the statistical window in the updated performance data table and the set hot spot threshold value, the hot spot caused by frequent operation of which type of transaction on which specific record can be accurately positioned, the positioning precision is high, and other unrelated services or unrelated records cannot be influenced by treatment based on the positioning precision. The hotspot analysis and positioning can be performed by receiving the transaction request packet and the transaction return packet and acquiring corresponding information, which is irrelevant to the internal architecture of a specific real-time processing system, so that the hotspot analysis and positioning method is generally applicable to various real-time processing systems and has strong universality. The hotspot analysis of the embodiment is specific to each transaction, when the hotspot record appears, the positioning can be rapidly completed, and the reaction speed can reach the second level by setting appropriate parameters, so that the reaction is rapidly reflected.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts. In the drawings:

FIG. 1 is a schematic flow chart of a method for monitoring and processing hot spot records in a real-time processing system according to an embodiment of the present invention;

FIG. 2 is a schematic flow chart of a method for monitoring and processing hot spot records in a real-time processing system according to another embodiment of the present invention;

FIG. 3 is a schematic flow chart of a method for monitoring and processing hot spot records in a real-time processing system according to another embodiment of the present invention;

FIG. 4 is a schematic flow chart diagram of a method for monitoring and processing hot spot records in a real-time processing system according to yet another embodiment of the present invention;

FIG. 5 is a schematic flow chart diagram of a method for monitoring and processing hot spot records in a real-time processing system according to yet another embodiment of the present invention;

FIG. 6 is a flow chart illustrating a method for updating a performance data table according to record information and a return timestamp in an embodiment of the present invention;

FIG. 7 is a flow chart illustrating a method for obtaining related records of information about transaction records in a performance data table according to an embodiment of the present invention;

FIG. 8 is a flow chart illustrating a method for obtaining related records of information about transaction records in a performance data table according to another embodiment of the present invention;

FIG. 9 is a flowchart illustrating a method for obtaining records related to transaction record information in a performance data table according to another embodiment of the present invention;

FIG. 10 is a flowchart illustrating a method for monitoring and processing hot spot records in a real-time processing system according to an embodiment of the present invention;

FIG. 11 is a diagram of more statistical information in an embodiment of the present invention;

FIG. 12 is a schematic diagram of a device for monitoring and processing hot spot records in a real-time processing system according to an embodiment of the present invention;

FIG. 13 is a schematic structural diagram of a device for monitoring and processing hot spot records in a real-time processing system according to another embodiment of the present invention;

FIG. 14 is a schematic structural diagram of a device for monitoring and processing hot spot records in a real-time processing system according to yet another embodiment of the present invention;

FIG. 15 is a schematic diagram of a hotspot record monitoring and processing device in a real-time processing system according to yet another embodiment of the invention;

FIG. 16 is a schematic diagram of a monitoring and processing device for hot spot records in a real-time processing system according to yet another embodiment of the present invention;

FIG. 17 is a block diagram of a performance data update unit according to an embodiment of the present invention;

fig. 18 is a schematic structural diagram of a registration-related field acquisition module according to an embodiment of the present invention;

FIG. 19 is a block diagram of a registration related field retrieving module according to another embodiment of the present invention;

FIG. 20 is a block diagram illustrating a structure of a registration-related field retrieving module according to another embodiment of the present invention;

FIG. 21 is a schematic structural diagram of a device for monitoring and processing hot spot records in a real-time processing system according to another embodiment of the present invention;

FIG. 22 is a diagram illustrating the slicing of time into a plurality of consecutive statistics windows according to the duration of the statistics windows in accordance with one embodiment of the present invention;

FIG. 23 is a schematic diagram of a real-time processing system with a hotspot record monitoring and processing device deployed outside of the real-time processing system in accordance with an embodiment of the present invention;

FIG. 24 is a schematic diagram of a real-time processing system with a hotspot record monitoring and processing device deployed therein according to an embodiment of the invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. The exemplary embodiments and descriptions of the present invention are provided to explain the present invention, but not to limit the present invention.

In the embodiments of the present invention, some words may be interpreted as the following meanings:

and (3) online processing: the information is directly input into the system from the generation place, and the processing result is directly sent to the place needing the information, which is also called on-line processing.

Real-time processing: the on-line processing is divided into real-time processing and delay processing. The former processes the input data immediately and obtains the result; the latter stores the input data and processes the data after a period of time.

Transaction: the transaction is a basic unit for data recovery and concurrency control, and has atomicity, consistency, isolation and durability. Typically corresponding to a complete transaction such as a transfer, deposit, remittance, etc.

Concurrent control: various techniques for protecting database integrity when multiple users are updating a runtime simultaneously. Locking data to be updated until the end of a transaction is the most common means for performing concurrency control and ensuring data integrity.

Recording hot spots: in a real-time processing system, when a large number of concurrent transactions need to update the same database record (such as the real-time cash-out meeting of a balance bank updates the account record of the balance bank in the spare payment account of a bank, the debt sale meeting of the balance bank updates the credit record of the balance bank, and the enterprise account record of the enterprise payment exchange meeting of the enterprise), the record is frequently applied for locking, which is called a hotspot record. The occurrence of hot spot records can cause transaction queuing and prolong response time, and because the resources locked by the transaction are various, the slow phenomenon is easy to quickly spread in the system, and if the processing is not timely, the whole real-time processing system can be paralyzed in a short time. Monitoring and timely disposal of hotspot records are very important for information systems of important industries and utilities, otherwise serious social impact and economic loss can be caused.

The method comprises the steps of obtaining relevant information of suspected hotspot records in a transaction request packet and a return packet through an inlet (outlet) for receiving transaction requests and returning transaction results in a real-time processing system, confirming specific records of current overheating and transaction request types causing hotspots through real-time statistical analysis according to the relevance between the hotspot records and the performance reduction of collective transactions, automatically triggering current limiting measures aiming at the transactions, and automatically removing current limiting when the hotspot access amount is reduced. The method is mainly realized based on the following two principles:

(1) the hot spot record related information can be obtained from a transaction request packet and a return packet of the real-time processing system, and comprises the following steps:

transaction performance data. The processing time of the transaction is used for representing the processing time, and the time interval between the return packet and the request packet can be calculated to obtain the processing time.

A transaction request type. Real-time processing systems typically receive multiple transaction requests, with the request differentiated by adding a transaction request code to the transaction request package to instruct the real-time processing system to take different processing modes, such as transfer, deposit, remittance, etc.

A specific record of the transaction. Which specific records the real-time processing system transaction is to include, either directly or indirectly, in the request package, such as the transfer transaction request package, or the customer number (if the real-time processing system agrees to use the customer's default account number).

(2) Slow correlation of hotspot records with collective processes:

hot spot logging can cause a decrease in the collective performance of related transactions on the log.

Sudden appearance of a certain transaction on a certain record with continuous (such as 60 seconds) and collective (such as 100 pens) performance anomalies can basically exclude other occasional interference factors, and the temporary hot spot phenomenon exists on the records. Current limiting is required to step down the real-time processing system.

Fig. 1 is a schematic flowchart of a method for monitoring and processing a hotspot record in a real-time processing system according to an embodiment of the present invention. As shown in fig. 1, a method for monitoring and processing a hotspot record in a real-time processing system according to an embodiment of the present invention may include:

s110: receiving a transaction request packet and recording a receiving time stamp of the transaction request packet;

s120: receiving a transaction return packet returned after the real-time processing system processes the transaction request packet, and recording a return timestamp of the transaction return packet;

s130: calculating the transaction response time according to the receiving timestamp and the returning timestamp;

s140: judging whether the transaction response time exceeds a set threshold processing time or not;

s150: if so, updating a performance data table according to the record information in the transaction request packet and the return timestamp, wherein the performance data table comprises a hotspot record information column corresponding to the record information, a statistic window ending time column and an abnormal transaction number column in a statistic window;

s160: judging whether a hot spot appears on the transaction type corresponding to the transaction request packet and the record information according to the updated data of the abnormal transaction number column in the statistical window in the performance data table and a set hot spot threshold value;

s170: and if so, limiting the transaction related to the transaction type and the record information according to set current limiting parameters.

In the step S110, the information of the received/intercepted transaction request packet may be temporarily stored in the memory for subsequent use. The receiving timestamp may be a time point t1 of receiving/intercepting the transaction request packet, and the time point t1 may be a time difference between the current time and 1900 year 1 month 1 day 00:00:00, and the unit may be milliseconds. For example, the timestamp at time point 02:00:00.001 on month 1 of year 2000 may be 3155767200001.

When the transaction processing is completed and returns from the real-time processing system in the above step S120, the transaction return packet may be received/intercepted, and the time point t2 at this time is recorded. Similar to the receive timestamp described above, time t2 may be the time difference between the time point and 1900 year 1 month 1 day 00:00:00, and may be in milliseconds.

In step S130, the transaction response time may represent a transaction processing time, which can reflect the speed of the current real-time processing system for processing the transaction. In step S140, the threshold processing time may be set by referring to the normal processing time of various transactions in the real-time processing system, and may be in milliseconds. For normally returned transactions (another part of transactions can be rejected by the current limiting module), a condition that the transaction processing time/transaction response time is larger than a set threshold processing time can be taken as a necessary condition for collecting the information related to the transactions.

In the step S150, if the transaction response time exceeds the set threshold processing time, relevant information of the transaction, such as specific record information and a return timestamp, may be collected. The collected information may be temporarily stored for hotspot analysis and positioning in steps S150 and S160, for example. By continuously updating the performance data table according to the record information and the return timestamp in the transaction request packet, the performance data table can be used for counting the transaction processing condition of the real-time processing system within a period of time, for example, counting the number of abnormal transactions within the latest period of time, so that the performance of the real-time processing system for processing the transactions can be reflected. The statistical time period may include a plurality of statistical windows, so the performance data table may include a number of abnormal transactions field in the plurality of statistical windows, which correspond to different statistical windows, respectively, the end time field of the statistical window may record an end time of the statistical window, and the end time of the statistical window may represent an end time of the statistical window at the latest time, that is, an end time of the entire statistical time period. The hot spot record information field can be found or established by using the record information in the transaction request packet, the data of the end time field of the statistical window can be updated to be the latest data by using the return timestamp, the transaction with the transaction response time exceeding the set threshold processing time can be considered as an abnormal transaction, and at the moment, the data in the abnormal transaction number field in the statistical window can be increased by one.

The performance data table may be pre-established. The performance data table may include records for a plurality of different types of transactions, and a transaction of the same type may include a plurality of different record information. In the embodiments of the present invention, the same record information of the same type of transaction is generally analyzed.

In step S160, by determining whether a hot spot occurs on the transaction type corresponding to the transaction request packet and the record information, it can be specifically located which record information of which transaction type the hot spot occurs on.

In the step S170, the transaction related to the transaction type and the recording information can be throttled according to which transaction type and which recording information transaction can be throttled, thereby avoiding a mischief.

In some embodiments, the transaction type is represented by data of a transaction type encoding field, and the transaction type encoding field is pre-added to the transaction request packet.

In some embodiments, the record information is obtained from an input packet start bit and an input packet end bit. The storage mode corresponding to the input packet start bit and the input packet end bit can save the storage space and can play a role in keeping the record information secret to a certain extent.

In the embodiment, the performance data table is updated according to the record information and the return timestamp in the transaction request packet, and whether a hot spot occurs on the transaction type and the record information corresponding to the transaction request packet is judged according to the data of the abnormal transaction number field in the statistical window in the updated performance data table and the set hot spot threshold value, so that the hot spot caused by frequent operation of which type of transaction on which specific record can be accurately positioned, the positioning precision is high, and other unrelated services or unrelated records cannot be influenced by treatment based on the hot spot. The hotspot analysis and positioning can be performed by receiving the transaction request packet and the transaction return packet and acquiring corresponding information, which is irrelevant to the internal architecture of a specific real-time processing system, so that the hotspot analysis and positioning method is generally applicable to various real-time processing systems and has strong universality. The hotspot analysis of the embodiment is specific to each transaction, when the hotspot record appears, the positioning can be rapidly completed, and the reaction speed can reach the second level by setting appropriate parameters, so that the reaction is rapidly reflected.

Fig. 2 is a schematic flow chart of a method for monitoring and processing a hotspot record in a real-time processing system according to another embodiment of the invention. As shown in fig. 2, the method for monitoring and processing a hotspot record in the real-time processing system shown in fig. 1 may further include, before step S120, that is, before receiving a transaction return packet returned after the real-time processing system processes the transaction request packet, the step of:

s180: judging whether the flow limitation needs to be carried out on the transaction corresponding to the transaction request packet or not according to the data of the abnormal transaction number column in the statistical window in the performance data table before updating and the set hot spot threshold value;

s190: if not, the transaction request packet is forwarded to the real-time processing system for logic processing.

In step S180, the performance data table before updating refers to the performance data table without analyzing the current transaction request packet processing situation, and may be, for example, the performance data table without updating in step S150. If the judgment result shows that the flow limitation of the transaction corresponding to the transaction request packet is not needed, the transaction request packet can be directly forwarded to the real-time processing system for logic processing.

In step S190, the transaction request packet may be forwarded to the real-time processing system for normal service logic processing.

In this embodiment, by determining whether the transaction needs to be limited before the real-time processing system processes the transaction request packet, the hot spot record can be processed in real time, the response is timely and rapid, and early warning analysis can be performed on the hot spot.

In some embodiments, the performance data table includes a plurality of abnormal transaction number fields in the statistical window, and the statistical window of the abnormal transaction number field in each statistical window is arranged in time sequence. In some embodiments, in step S180, the method for determining whether the transaction corresponding to the transaction request packet needs to be limited according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot spot threshold may be implemented as follows: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the performance data table before updating exceeds the set hot spot threshold value, if so, determining that the current limitation needs to be performed on the transaction corresponding to the transaction request packet.

Fig. 3 is a flowchart illustrating a method for monitoring and processing a hotspot record in a real-time processing system according to another embodiment of the invention. As shown in fig. 3, before step S180, that is, before determining whether to perform current limiting on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot spot threshold, the method for monitoring and processing the hot spot record in the real-time processing system shown in fig. 2 may further include the steps of:

s1100: judging whether a performance data record related to the record information exists in the performance data by searching the performance data table;

s1110: if not, the step of forwarding the transaction request packet to the real-time processing system for logic processing is executed; if yes, judging whether the performance data record of the record information meets the following conditions: and if not, forwarding the transaction request packet to the real-time processing system for logic processing, and if so, judging whether the flow limitation needs to be performed on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot point threshold value.

In the step S110, if the performance data table determines that there is no performance data record related to the record information in the transaction request packet, which indicates that performance monitoring is not required, the transaction request packet may be directly forwarded to the real-time processing system for logic processing. If the performance data table determines that there is a performance data record related to the record information in the transaction request packet, it may continue to check whether the performance data record is valid, and the determination method may be: the end time of the statistical window > current time point-statistical window duration x number of statistical windows, wherein the current time point may refer to a time point at which a transaction is received. The data of the statistic window ending time field in the performance data table may be the statistic window ending time stored in the current performance data table. The current time point may be a time point of the time, for example, a time point of the judgment time point may be a time difference between the time point and 1 month 1 day 00:00:00 of 1900 year, and the unit may be milliseconds. If the performance data record is judged to be invalid, the current time point is not in the statistical window, abnormal data are not counted nearby the time point of the transaction, and current limitation is not needed, the transaction request packet can be directly forwarded to the real-time processing system for logic processing, the transaction is not counted in the statistical range, and therefore statistics of the transaction in the statistical window can be guaranteed. If the performance data record is judged to be valid, a hot spot threshold value (current limiting condition) is set to judge whether to limit the current.

Fig. 4 is a flowchart illustrating a method for monitoring and processing a hot spot record in a real-time processing system according to still another embodiment of the present invention. As shown in fig. 4, the method for monitoring and processing hot spot records in the real-time processing system shown in fig. 2 may further include the steps of:

s1120: if the transaction corresponding to the transaction request packet needs to be limited, judging whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in a current limiting window corresponding to the record information in the performance data table before updating;

s1130: if yes, generating a transaction rejection packet according to the transaction request packet, taking the time point when the transaction rejection packet is generated as a return timestamp, and executing the step of updating a performance data table according to the record information in the transaction request packet and the return timestamp; if not, updating the data of the forwarded transaction number column in the current limiting window corresponding to the record information in the performance data table, and executing the step of forwarding the transaction request packet to the real-time processing system for logic processing.

In step S1130, the transaction that needs to be limited may be generated into a return packet with a special format, and the reject packet may be a special form of the return packet, and may also notify the receiving party that the system is busy and the request is not accepted. If the real-time processing system needs to be prevented from processing the transaction request packet, after a transaction rejection packet is generated, a step of updating the performance data table according to the record information in the transaction request packet and the return timestamp (step S150) can be executed, and the transaction corresponding to the transaction rejection packet can be regarded as an abnormal transaction, so that the data of the abnormal transaction number field in the statistical window in the performance data table can be updated, and thus the transaction rejection packet can also be used for hotspot analysis and positioning, and the analysis result is more accurate. In step S150, the return timestamp may be replaced by a time point when the transaction rejection packet is generated, and the record information in the transaction request packet may be obtained from the transaction rejection packet or from the temporary information when the transaction request packet is accepted. The reject packet may be returned to the location in the online process where the information is needed. If the transaction request packet does not need to be prevented from being processed by the real-time processing system, the data in the forwarded transaction number field in the current limiting window corresponding to the record information in the performance data table may be updated, for example, incremented by one, and the step of forwarding the transaction request packet to the real-time processing system for logic processing is performed (step S190).

In some embodiments, the setting the current limit parameter may include: the time length of the current limiting window, the number of the current limiting windows and the number of transactions allowed to be forwarded by each current limiting window. The performance data table may further include a number of forwarded transactions field in the current limiting window, where the number of forwarded transactions field in each current limiting window is equal to the number of current limiting windows, and the current limiting windows in the forwarded transactions field in each current limiting window have the duration of the current limiting window and are arranged in time sequence. In the step S1120, the method for determining whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in the current limiting window corresponding to the record information in the performance data table before updating may be implemented as follows: and judging whether the real-time processing system needs to be prevented from processing the transaction request packet or not according to whether the data of the forwarded transaction number field in each current limiting window corresponding to the record information in the performance data table before updating is larger than or equal to the number of the transactions allowed to be forwarded by the current limiting window.

In this embodiment, it may be checked whether "the number of forwarded transactions in the current limiting window" reaches the number of current limiting transactions, and the determining method includes: and the number of forwarded transactions in the current limiting window is equal to the number of current limiting transactions. And updating the data of the forwarded transaction number column in each current-limiting window according to the judgment result of each current-limiting window. In this embodiment, by setting a plurality of current limiting windows, uniform current limiting can be achieved, and a better current limiting effect can be obtained.

Fig. 5 is a flowchart illustrating a method for monitoring and processing a hot spot record in a real-time processing system according to still another embodiment of the present invention. As shown in fig. 5, before step S1120, that is, before determining whether it is necessary to prevent the real-time processing system from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in the current limiting window corresponding to the record information in the performance data table before updating, the method for monitoring and processing the hotspot record in the real-time processing system shown in fig. 4 may further include the steps of:

s1140: and judging whether the current time point is less than or equal to the data of the current limiting window ending time column corresponding to the record information in the performance data table, if not, updating the data of the current limiting window ending time column according to the current time point and the set current limiting window duration.

In step S1140, the current time point may be a time point of receiving the transaction, and may be a time difference between the current time point and 1 month, 1900 and 1 month, 00:00:00, and the unit may be milliseconds. When the current time point is less than or equal to the data of the current limiting window ending time field corresponding to the record information in the performance data table, if the number of the forwarded transactions in the current limiting window does not reach the number of the current limiting transactions, the number of the forwarded transactions in the current limiting window can be updated, for example, the data of the number of the forwarded transactions in the current limiting window is increased by one. When the current time point is greater than the data of the current-limiting-window ending-time field corresponding to the record information in the performance data table, the data of the current-limiting-window ending-time field may be updated according to the current time point and the set current-limiting-window duration, and the specific method includes: the end time of the current limiting window is (rounded (current time point/duration of the current limiting window) +1) × the duration of the current limiting window, that is, the data of the end time field of the current limiting window is (rounded (current time point/duration of the set current limiting window) +1) × the duration of the set current limiting window.

In this embodiment, by determining whether the current time point is less than or equal to the data of the current-limiting window ending time column corresponding to the record information in the performance data table, it may be checked whether the current time point is in the current-limiting window in the performance data record, that is, whether the following conditions are satisfied: current limit window end time > current point in time.

In some embodiments, the performance data table includes a plurality of abnormal transaction number fields in the statistical window, and the statistical window of the abnormal transaction number field in each statistical window is arranged in time sequence. In this embodiment, the time may be divided into a plurality of continuous windows according to the value specified by the "statistical window duration" parameter, so that the statistical data in the performance data table may better conform to the actual processing condition of the real-time processing system, and the accuracy of hotspot positioning and current limiting may be further improved.

In some embodiments, in step S160, a method for determining whether a hot spot occurs in the transaction type corresponding to the transaction request packet and the record information according to the updated data of the abnormal transaction number field in the statistical window in the performance data table and a set hot spot threshold may be implemented as follows: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the updated performance data table exceeds the set hot spot threshold value, and if so, confirming that a hot spot appears on the transaction type and the record information.

FIG. 6 is a flow chart illustrating a method for updating a performance data table according to record information and a return timestamp in an embodiment of the invention. As shown in fig. 6, in step S150, the method for updating the performance data table according to the record information in the transaction request packet and the return timestamp may include the steps of:

s151: acquiring a statistical window ending time column and an abnormal transaction quantity column in a statistical window related to the record information in the performance data table;

s152: calculating new statistical window ending time according to the return timestamp and the set statistical window duration so that the return timestamp is smaller than the new statistical window ending time, and updating the data of the acquired statistical window ending time column by using the new statistical window ending time;

s153: calculating the new end time of the statistical window of the abnormal transaction quantity column in each acquired statistical window according to the new end time of the statistical window and the set statistical window duration; comparing whether the new end time is the same as the window end time corresponding to the abnormal transaction number column in the obtained statistical window or not; if the two end times are the same, updating the data of the abnormal transaction number field in the obtained statistical window corresponding to the window end time by using the abnormal transaction number field in the obtained statistical window corresponding to the new end time; and adding one to the data of the abnormal transaction number field in the statistical window acquired at the last time, and updating the data of the abnormal transaction number field in the acquired statistical window which is not updated to zero.

In this embodiment, the end time of the statistical window (the end time of the statistical total period/the end time of the latest statistical window) can be continuously updated through the step S152, and the end time of each statistical window can be further continuously updated through the step S153. In step S153, the data in the field of the number of abnormal transactions in the statistical window in the performance data table may be updated according to the updated end time (new end time) of each statistical window. And when the transaction of the transaction request packet processing is considered to belong to the abnormal transaction, adding one to the data of the abnormal transaction number field in the statistical window acquired at the last time. Therefore, the statistical data in the performance data table can be ensured to be the data which is the nearest to the processing time of the request packet, so that the transaction processing performance condition of the real-time processing system can be reflected better in real time.

In one embodiment, in step S152, a new end time of the statistical window may be calculated according to the set statistical window duration "(rounded (return timestamp/set statistical window duration) + 1).

FIG. 7 is a flowchart illustrating a method for obtaining related records of information about transaction records in a performance data table according to an embodiment of the present invention. As shown in fig. 7, in the step S151, the method for obtaining the end time field of the statistic window and the number field of abnormal transactions in the statistic window related to the record information in the performance data table may include the steps of:

s1511: judging whether the performance data table contains a hotspot recording information column corresponding to the recording information or not by searching the performance data table;

s1512: if yes, obtaining a statistical window ending time field and an abnormal transaction quantity field in the statistical window related to the record information according to the searched hotspot record information field.

In the above step S1511 and step S1512, for example, the performance data table may be searched using information such as configuration codes and record information of the transactions, and if records of the same contents such as configuration codes and record information are searched in the performance data table, the search may be completed, and the end time field of the statistical window and the abnormal number of transactions in the statistical window related to the record information may be obtained according to the searched hotspot record information field.

In this embodiment, the fields related to the record information in the performance data table can be searched and obtained, so that the data of the related fields in the performance data table can be directly updated.

Fig. 8 is a flowchart illustrating a method for obtaining related records of information about transaction records in a performance data table according to another embodiment of the present invention. As shown in fig. 8, the method for obtaining the end time field of the statistic window and the number of abnormal transactions in the statistic window related to the logging information in the performance data table shown in fig. 7 may further include the steps of:

s1513: if the performance data table does not contain the hotspot record information column corresponding to the record information, adding the hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window related to the record information in the performance data table, and initializing the added hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window.

In the step S1513, the added data of the hot spot log information field may be the log information, and the end time field of the counting window and the number of abnormal transactions field in the counting window may be initialized to zero, for example.

In this embodiment, a new performance data record may be added to the performance data table, so that performance record data may be added to a transaction that has not been recorded in the performance data table before.

Fig. 9 is a flowchart illustrating a method for obtaining related records of information about transaction records in a performance data table according to another embodiment of the present invention. As shown in fig. 9, the method for obtaining the end time field of the statistic window and the number of abnormal transactions in the statistic window related to the logging information in the performance data table shown in fig. 7 may further include the steps of:

s1514: if the performance data table does not contain the hotspot record information column corresponding to the record information, searching and finding a statistic window failure record in the performance data table, initializing the statistic window failure record, and respectively taking the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the initialized statistic window failure record as the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the record information, wherein the statistic window failure record satisfies the following conditions: and returning a timestamp > the end time of the statistical window in the statistical window failure record + the number of the set statistical windows and the set statistical window duration.

In step S1514, the record in the performance data table satisfying the following condition may be a statistical window failure record: and returning a time point > the end time of the statistical window + the number of the statistical windows x the duration of the statistical windows, if the performance data table does not have the performance data record of the record information, searching the performance data table to find out the record of failure of the first statistical window, wherein the record in the performance data table can be recovered and reused, so that the waste of the storage space of the system can be avoided, and the storage pressure of the system is reduced.

Fig. 10 is a flowchart illustrating a method for monitoring and processing a hotspot record in a real-time processing system according to an embodiment of the invention. As shown in fig. 10, the method for monitoring and processing hot spot records in the real-time processing system of this embodiment may include one or more of the following steps:

(a1) configuring acquisition conditions

The normal processing time of each transaction can be known through the analysis of the transaction processing logic. The configured acquisition conditions may include configuration code, transaction type, threshold processing time, etc. The fields associated with the acquisition conditions in the configuration table can be as shown in table 1.

Table 1 configuration table fields associated with acquisition conditions

The information about the acquisition conditions shown in table 1 means that the acquisition is required for any one of the records:

(1) the transaction type is "10011" (treasury sales transaction) and the processing time exceeds 2000 milliseconds.

(2) The transaction type is "10022" (a surreptitious payroll transaction) and the processing time exceeds 1000 milliseconds.

(a2) Configuring acquisition content

By analyzing the transaction processing logic, it can be known which types of records need to be locked by each transaction, and usually the transaction request packet includes specific information (such as key values, or information related to key values) for locating these records so that the transaction can be correctly processed. For example, a payroll transaction requires locking of the transfer-out account and the transfer-in account. And the transaction request packet contains the related information of the transferred account and the transferred account. It should be noted that all locked records are not easy to generate hotspots, and only the record information easy to generate hotspots needs to be collected and monitored, for example, a payroll transaction is usually the transfer of one account to a plurality of accounts, a roll-out account is locked for a plurality of times and is easy to generate hotspots, a roll-in account is only locked once and is not easy to generate hotspots, and only the roll-out account information is collected and monitored. The configured acquisition content may include one or more pairs of input packet start bits and input packet end bits. The fields associated with the acquisition content in the configuration table can be as shown in table 2.

Table 2 field of configuration table related to the collected content

The information related to the collected content shown in table 2 means that when the screening condition is satisfied, the following content is collected and transmitted to a subsequent analysis step for use:

(1) the transaction type is '10011' (national bond sales transaction) and the processing time exceeds 2000 milliseconds, the 5 th to 15 th contents (national bond credit account) of the transaction request packet need to be collected, and the corresponding configuration code is '001'.

(2) The transaction type is '10022' (pay for generation transaction) and the processing time exceeds 1000 milliseconds, the contents of 10 th to 29 th (transfer deposit primary account number) and 50 th to 53 th (transfer deposit sub account number) of the transaction request packet need to be collected, and the corresponding configuration code is '002'.

(a3) Configuring hot spot judgment conditions

The number of abnormal transaction processing performances in the latest time period (for example, the processing time exceeds a threshold) can be counted according to the collected information, and the records of which the transactions are processed too slowly in a collective manner can be analyzed, other occasional interference factors can be basically eliminated when the transactions are processed too slowly in the collective manner, and the temporary hot spot phenomenon existing on the records is positioned, so that measures are required to be taken, and the system safety is ensured. The configured hotspot judgment condition may be that the number of transaction performance anomalies in the latest statistical time period exceeds a set hotspot threshold. The fields associated with alarm conditions in the configuration table may be as shown in table 3.

TABLE 3 field of configuration table associated with alarm condition

The hot spot determination condition information in table 3 means that a hot spot occurs when any of the following conditions is satisfied:

(1) the transaction type is '10011' (national bond sales transaction) and the processing time exceeds 2000 milliseconds, and the total number in the latest 12 5000 milli-second time windows is more than 30 according to the statistical number of the content groups from the 5 th bit to the 15 th bit (namely the content of the 5 th bit to the 15 th bit is the same) of the transaction request packet, so that an alarm can be given. The transaction "10011" (treasury sale) is prompted for a hotspot on the treasury line account (5 th to 15 th of the transaction request package).

(2) The transaction type is '10022' (payroll transaction) and the processing time exceeds 1000 milliseconds, and according to the statistical number of the content groups from 10 th bit to 29 th bit and from 50 th bit to 53 th bit (namely, the content of the 10 th bit to the 29 th bit and the content of the 50 th bit to the 53 th bit are all the same) of the transaction request packet, the total number in the latest 12 5000 millisecond time windows is more than 30, and then an alarm can be given. The transaction "10022" (commission) is prompted for a hotspot on the transfer-out deposit sub-account number (10 th to 29 th, 50 th to 53 th of the transaction request packet).

(a4) Configuring current limiting parameters

The time may be divided into a plurality of consecutive windows according to the value specified by the "time duration of the current limiting window" parameter, starting at 1900 s 1/00: 00: 00.000. When a hot spot is identified, the flow can be limited by setting the duration of the flow limiting window, and the number of transactions allowed to pass within each window. The configured current limit parameters may include information such as a current limit window duration, a number of current limit transactions, and the like. The fields associated with the current limit condition in the configuration table may be as shown in table 4.

Table 4 field of configuration table associated with current limit condition

The current limiting parameter information shown in table 4 means that the following current limiting actions are taken after identifying a hot spot:

(1) in the first case, a current limit window duration of 0 may indicate that no current limit action is taken.

(2) In the second case, the transaction type is "10022" (payroll transaction) on the identified hotspot record, only 100 passes are allowed in each 5000 millisecond time window, and more than 100 immediately can return a reject packet.

In some embodiments, as shown in FIG. 10, the following steps are included when processing the transaction:

(b1) receiving a transaction request packet

Before the transaction request reaches the real-time processing system, the request packet information can be intercepted.

(b2) Temporarily storing request packet information and recording time stamp

The intercepted request packet information may be temporarily stored in the memory, and the time point t1 at this time is recorded, where the time point t1 may be a time difference between the current time point and 1900 year 1 month 1 day 00:00:00, and the unit is milliseconds, for example, 3155767200001 is t1 at the time point of 2000 year 1 month 1 day 02:00: 00.001.

The request packet may then be forwarded to the current limit module.

(b3) Judging whether to limit the current

The flow limiting module receives the transaction uplink packet, checks the configuration table and the performance data table, and can execute the following steps:

(b3-1) intercepting the transaction type and hot spot record information from the transaction upper package, and searching the performance data record in the performance data table. If executable is found (b3-2), otherwise executable (b 4).

(b3-2) checking whether the performance data record is valid, i.e. the end time of the statistical window was recently updated due to the occurrence of an abnormal transaction. The judging method can be as follows: end time of the statistical window 1 > current time point-statistical window duration x number of statistical windows. If so, it may be executed (b3-3), otherwise, it may be executed (b 4).

(b3-3) checking whether the hot spot judgment condition is satisfied, and the judgment method can be: the sum of the number of abnormal transactions in all the statistical windows is greater than the hotspot threshold. If so, it may be executed (b3-4), otherwise, it may be executed (b 4).

(b3-4) checking whether the current time point is within the current limiting window in the performance data record, and the determining method can be: current limit window end time > current point in time. If so, it may be executed (b3-5), otherwise, it may be executed (b 3-7).

(b3-5) checking whether the number of the forwarded transactions in the current limiting window reaches the number of the current limiting transactions, wherein the judging method comprises the following steps: and the number of forwarded transactions in the current limiting window is equal to the number of current limiting transactions. If so, it may be executed (b5), otherwise, it may be executed (b 3-6).

(b3-6) updating the "number of forwarded transactions within the current limit window". The updating method can be as follows: the value of "number of forwarded transactions within the current limit window" is incremented by 1. Then (b4) may be performed.

(b3-7) updating the current limit window ending time and the forwarded transaction number in the current limit window, wherein the updating method can be as follows: the end time of the current limiting window is (round (current time point/current limiting window duration) +1) the duration of the current limiting window; the value of "number of forwarded transactions within the current limit window" is 1. Then (b4) may be performed.

(b4) Forwarding transaction requests

And normally forwarding the transaction request packet to a real-time processing system for service processing.

(b5) Generating transaction rejection packages

Transactions that require throttling generate return packets in a special format. The reject packet should be a special form of return packet, informing the recipient that the system is busy and the request is not accepted.

(b6) Accepting transaction return packet record timestamps

When the transaction is completed and returns from the real-time processing system, the return packet is intercepted and the time point t2 at this time is recorded. Time t2 may be the difference in milliseconds between the time and day 1 of 1900, 1 month 00:00:00.

(b7) Judging whether to collect

Checking the return packet or the reject packet, and according to the configuration table, judging whether to collect according to the following steps:

(b7-1) if it is a reject packet, then (b8) may be executed, otherwise (b7-2) may be executed.

(b7-2) calculating the processing time of the transaction through the time stamps recorded in the steps (b6) and (b1), wherein the processing time t is t2-t 1. If the transaction processing time t is larger than the corresponding threshold processing time in the configuration table; and the transaction type in the upload package is in the configuration table, then it may be executed (b8), otherwise it may be executed (b 9).

(b8) Collecting

Based on the configuration table (which may be loaded into memory for efficiency), the following information is collected and may be transmitted to subsequent analysis steps in an asynchronous manner (e.g., via queue techniques such as MQ). And an asynchronous processing mode is adopted, so that the influence on the transaction processing request is very small, and the transaction response time is not basically increased. The configuration table may be as shown in table 5.

TABLE 5 Collection of configuration Table

The configuration code can be used for uniquely identifying a piece of configuration information and a key value of a configuration table, and is not repeatable. The hotspot recording information can intercept the content in the transaction request packet according to the input packet start bit and the input packet end bit in the configuration table, and when a plurality of pairs of input packet start bit and input packet end bit information exist in the configuration table, the plurality of pairs are spliced according to the sequence, so that the storage space can be saved, and the confidentiality is increased. Returning to the point in time, it may be the point in time t2 recorded in step (b 6).

(b9) Forwarding transaction return packets

The intercepted transaction return packet or transaction reject packet may be forwarded to the system submitting the transaction request to complete the transaction return.

An asynchronous hanging mode can be adopted, and collected information and hotspot positioning records are analyzed:

(b10) analysis, hotspot record location

And (4) asynchronously reading the information (configuration codes and hotspot record information) transmitted in the step (b8) one by one in a frequently hanging manner, and calculating and updating the performance data table. And an asynchronous processing mode is adopted, so that the influence on the transaction processing request is very small, and the transaction response time is not basically increased. The treatment can be specifically carried out circularly according to the following steps:

(b10-1) locating performance data records: and (4) searching a performance data table by using the configuration code and the hotspot record information, if the records with the same configuration code and hotspot record information content are searched, finishing positioning, and jumping to the step (b10-10), otherwise, performing the step (b 10-2).

(b10-2) recovery Performance data record: and searching the performance data table to find the record of the failure of the first statistical window. The record in the performance data table that satisfies the following conditions may be a statistical window failure record:

return time point > end time of the statistical window + number of statistical windows.

After a record with a failed statistical window is successfully searched, the record can be recovered, and fields of the record can be initialized according to the method shown in table 6 below.

TABLE 6 initialization results for each field in the recycle record

If the performance data records are successfully reclaimed, then step (b10-10) can be skipped, otherwise step (b10-3) can be continued.

(b10-3) adding performance data records: a record is newly added to the performance data table. The initial value of each field of the newly added record can be shown in table 7.

TABLE 7 initial value of newly added record of each field

(b10-10) calculating the end time of the current transaction statistics window. The end time of the current transaction statistics window may be denoted as m. For example, the end time m of the current transaction statistics window is (rounded (t 2/statistics window duration) +1) × statistics window duration. The statistical window duration may be from the configuration table described above.

(b10-11) calculating the end time of each current statistical window. The end time of each current statistical window can be recorded as m1, m2, … …, respectively. Then there may be:

m1＝m

m 2-m 1-statistical window duration

m 3-m 2-statistical window duration

……

The statistical window duration and the statistical window number can be from the configuration table.

(b10-12) calculating the end time of the original statistical window. The end times of the original statistical windows can be denoted as w1, w2, … …, respectively. Then there may be:

w1 is the value of the field "end time of statistical Window" in the statistical record

w2 ═ w 1-statistical window duration

w3 ═ w 2-statistical window duration

The time length of the statistical window and the number of the statistical windows are from the configuration table.

(b10-13) updating the statistical information. In one embodiment, the number of abnormal transactions in the statistical window 1, … …, and the number of abnormal transactions in the statistical window 1 may be updated as follows: comparing w1, w2, w3 … … with m1, m2 and m3 … …, and assigning the value of the 'number of abnormal transactions in the statistical window j' to the 'number of abnormal transactions in the statistical window i' field when mi is wj; the number of abnormal transactions in other statistical windows is assigned to be 0; setting the end time of the statistical window as the end time m of the current transaction statistical window; the number of anomalous transactions within statistical window 1 may be increased by 1.

FIG. 11 is a diagram illustrating further statistics in an embodiment of the present invention. As shown in fig. 11, the end time of the original statistical window may be updated by using the end time of the current statistical window (new statistical window end time), and the end time of each statistical window may be updated by using the statistical window duration (5 s). When an abnormal transaction occurs at the transaction return time point/return timestamp, the data of the abnormal transaction number field in the statistical window at the latest statistical window after updating (the statistical window at the transaction return time point) may be increased by one. The content of step S153 described above can be better understood from fig. 11.

When the sum of the number of abnormal transactions in all the statistical windows is larger than the hot spot threshold value, hot spot records can be shown to appear. The updated data of this step may be used as input to step (b3) for triggering current limiting.

The steps (a1), (a2), (a3), and (a4) are preparatory steps for configuring parameters. And a method for flexibly acquiring the related information of the hotspot record in a configuration mode from the transaction request packet and the transaction return packet. The hot spot record related information comprises transaction processing time, transaction types and specific information capable of identifying records, and the corresponding data acquisition module can comprise a step (b1), a step (b2), a step (b6), a step (b7), a step (b8) and a step (b 9). By utilizing the correlation between the hotspot records and the slow collective processing, the specific hotspot records can be found by a statistical method. Hotspot records can cause the collective performance of related transactions on the record to degrade; the sudden appearance of a certain transaction on a certain record (for example, 60 seconds), the clustering performance degradation can basically exclude other occasional interference factors, and the record is positioned to have a temporary hot spot phenomenon, and the corresponding analysis module can comprise a step (b 10). The asynchronous processing mode of the analysis module can ensure that the real-time processing system has little influence on the transaction processing, and the analysis module can comprise the step (b 10). The method of slicing time is adopted to form continuous time windows, the transaction condition is monitored by a statistical method for a plurality of recent continuous time windows, the analyzed data is ensured to continuously reflect the latest state in real time, and the corresponding analysis module can comprise the step (b 10). The current limiting has good accuracy for specific transaction types and specific hot spot records, and the corresponding hot spot current limiting module may include steps (b3), (b4), and (b 5). The current limiting method adopts a total method in a time window, can effectively utilize the buffering capacity of a real-time processing system to transactions under the condition of not causing congestion of the real-time processing system, processes the transactions as much as possible, corresponds to a hot spot current limiting module, and comprises the steps of (b3), (b4) and (b 5). The whole process can be automated, real-time reactive, configurable, corresponding to the configuration data module, and can include the steps (a1), (a2), (a3), (a 4).

The method of the embodiment can quickly and accurately position the hot spot record appearing in the real-time processing system; automatically and accurately limiting the current of the request transaction causing the hotspot record, and cooling the hotspot; and when the transaction request for operating the hotspot record is reduced, the current limiting measure is automatically removed, so that the safe and stable operation of the system is ensured. The method has the characteristics of quick response, accurate positioning, automatic disposal, convenient modification, strong universality and high performance, plays an important role in ensuring the smooth operation of large-scale important real-time processing systems (such as banks, securities and public utilities) and preventing the system from being blocked or even collapsing, and has good social benefit and economic benefit. When the method is adopted in a real-time processing system, when the hotspot record appears, the positioning can be rapidly completed, the rapid response can be realized, and the response speed can reach the second level according to the parameter setting; the method can accurately position which type of affairs cause hot spots due to frequent operation on which specific record, has high positioning precision, and can not influence other irrelevant services or irrelevant records by treatment based on the hot spots, thereby achieving the aim of accurate positioning; by supporting automatic current limiting, the reaction speed can reach the second level, and the current limiting measures are automatically removed after the transaction request quantity is reduced, so that automatic disposal is realized, and the automatic reaction speed of the second level can ensure safe and stable operation of an important information system because the hotspot recording spreading speed is very high; the method is generally applicable to various real-time processing systems, has nothing to do with the internal architecture of the specific real-time processing system, and has strong universality. The analysis and early warning module adopts an asynchronous processing mode, so that the influence on the transaction processing request is very small, the transaction response time is not basically increased, and the performance is high.

According to the monitoring and processing method of the hotspot record in the real-time processing system, the performance data table is updated according to the record information and the return timestamp in the transaction request packet, and whether the hotspot occurs on the transaction type and the record information corresponding to the transaction request packet is judged according to the data of the abnormal transaction number field in the statistical window in the updated performance data table and the set hotspot threshold value, so that the hotspot caused by frequent operation of which type of transaction on which specific record can be accurately positioned, the positioning precision is high, and the treatment performed based on the hotspot can not influence other unrelated services or unrelated records. The hotspot analysis and positioning can be performed by receiving the transaction request packet and the transaction return packet and acquiring corresponding information, which is irrelevant to the internal architecture of a specific real-time processing system, so that the hotspot analysis and positioning method is generally applicable to various real-time processing systems and has strong universality. The hotspot analysis of the embodiment is specific to each transaction, when the hotspot record appears, the positioning can be rapidly completed, and the reaction speed can reach the second level by setting appropriate parameters, so that the reaction is rapidly reflected.

Based on the same inventive concept as the monitoring and processing method for hotspot records in the real-time processing system shown in fig. 1, the embodiment of the present application further provides a monitoring and processing device for hotspot records in the real-time processing system, as described in the following embodiments. Because the principle of solving the problems of the monitoring and processing device for the hot spot record in the real-time processing system is similar to the monitoring and processing method for the hot spot record in the real-time processing system, the implementation of the monitoring and processing device for the hot spot record in the real-time processing system can refer to the implementation of the monitoring and processing method for the hot spot record in the real-time processing system, and repeated parts are not described again.

Fig. 12 is a schematic diagram of a device for monitoring and processing hot spot records in a real-time processing system according to an embodiment of the present invention. As shown in fig. 12, the monitoring and processing device for hot spot records in the real-time processing system of this embodiment may include: a request packet capturing unit 210, a return packet capturing unit 220, a response time calculating unit 230, an acquired information judging unit 240, a performance data updating unit 250, a hotspot record judging unit 260 and a hotspot record processing unit 270, which are connected in sequence.

The request packet intercepting unit 210 is configured to: and receiving a transaction request packet, and recording a receiving time stamp of the transaction request packet.

The return packet interception unit 220 is configured to: and receiving a transaction return packet returned after the real-time processing system processes the transaction request packet, and recording a return timestamp of the transaction return packet.

The response time calculation unit 230 is configured to: and calculating the transaction response time according to the receiving timestamp and the returning timestamp.

The collected information judgment unit 240 is configured to: and judging whether the transaction response time exceeds a set threshold processing time.

The performance data updating unit 250 is configured to: and if so, updating a performance data table according to the record information in the transaction request packet and the return timestamp, wherein the performance data table comprises a hotspot record information column corresponding to the record information, a statistic window ending time column and an abnormal transaction number column in the statistic window.

The hot spot record determining unit 260 is configured to: and judging whether hot spots appear on the transaction type corresponding to the transaction request packet and the record information according to the updated data of the abnormal transaction number field in the statistical window in the performance data table and a set hot spot threshold value.

The hotspot record processing unit 270 is configured to: and if so, limiting the transaction related to the transaction type and the record information according to set current limiting parameters.

In the request packet intercepting unit 210, the information of the received/intercepted transaction request packet may be temporarily stored in the memory for subsequent use. The receiving timestamp may be a time point t1 of receiving/intercepting the transaction request packet, and the time point t1 may be a time difference between the current time and 1900 year 1 month 1 day 00:00:00, and the unit may be milliseconds. For example, the timestamp at time point 02:00:00.001 on month 1 of year 2000 may be 3155767200001.

In the above-described return packet intercepting unit 220, when the transaction processing is completed and returns from the real-time processing system, the transaction return packet may be received/intercepted, and the time point t2 at this time is recorded. Similar to the receive timestamp described above, time t2 may be the time difference between the time point and 1900 year 1 month 1 day 00:00:00, and may be in milliseconds.

In the response time calculation unit 230, the transaction response time may represent a transaction processing time, which can reflect the speed of the current real-time processing system for processing the transaction. In the collected information determining unit 240, the processing time of the set threshold may be set according to the normal processing time of various transactions of the real-time processing system, and the unit may be milliseconds. For normally returned transactions (another part of transactions can be rejected by the current limiting module), a condition that the transaction processing time/transaction response time is larger than a set threshold processing time can be taken as a necessary condition for collecting the information related to the transactions.

In the performance data updating unit 250, if the transaction response time exceeds a set threshold processing time, relevant information of the transaction, such as specific record information, a return timestamp, and the like, may be collected. The collected information may be temporarily stored for use in, for example, performing hotspot analysis and positioning in the performance data updating unit 250 and the hotspot record judging unit 260. By continuously updating the performance data table according to the record information and the return timestamp in the transaction request packet, the performance data table can be used for counting the transaction processing condition of the real-time processing system within a period of time, for example, counting the number of abnormal transactions within the latest period of time, so that the performance of the real-time processing system for processing the transactions can be reflected. The statistical time period may include a plurality of statistical windows, so the performance data table may include a number of abnormal transactions field in the plurality of statistical windows, which correspond to different statistical windows, respectively, the end time field of the statistical window may record an end time of the statistical window, and the end time of the statistical window may represent an end time of the statistical window at the latest time, that is, an end time of the entire statistical time period. The hot spot record information field can be found or established by using the record information in the transaction request packet, the data of the end time field of the statistical window can be updated to be the latest data by using the return timestamp, the transaction with the transaction response time exceeding the set threshold processing time can be considered as an abnormal transaction, and at the moment, the data in the abnormal transaction number field in the statistical window can be increased by one.

In the hot spot record determining unit 260, by determining whether a hot spot occurs on the transaction type corresponding to the transaction request packet and the record information, it may be specifically located on which record information of which transaction type the hot spot occurs.

In the hot spot record processing unit 270, the transaction related to the transaction type and the record information is throttled, so that the transaction related to the record information of which specific transaction type can be throttled, thereby avoiding a clever and innocent affairs.

In some embodiments, the performance data updating unit 250 may further be configured to: the transaction type is represented by data of a transaction type encoding field, which is pre-added to the transaction request packet.

In some embodiments, the performance data updating unit 250 may further be configured to: and the recording information is obtained according to the start bit and the end bit of the input packet. The storage mode corresponding to the input packet start bit and the input packet end bit can save the storage space and can play a role in keeping the record information secret to a certain extent.

Fig. 13 is a schematic structural diagram of a monitoring and processing device for hot spot records in a real-time processing system according to another embodiment of the present invention. As shown in fig. 13, the monitoring and processing device for hot spot recording in the real-time processing system shown in fig. 12 may further include: a transaction flow limit judging unit 280 and a request packet forwarding unit 290, which are connected to each other. The transaction flow limit determining unit 280 is connected to the request packet capturing unit 210, and the request packet forwarding unit 290 is connected to the return packet capturing unit 220.

The transaction flow limit determining unit 280 is configured to: and judging whether the flow limitation needs to be carried out on the transaction corresponding to the transaction request packet or not according to the data of the abnormal transaction number column in the statistical window in the performance data table before updating and the set hot spot threshold value.

The request packet forwarding unit 290 is configured to: if not, the transaction request packet is forwarded to the real-time processing system for logic processing.

In some embodiments, the performance data table includes a plurality of abnormal transaction number fields in the statistical window, and the statistical window of the abnormal transaction number field in each statistical window is arranged in time sequence. In some embodiments, the transaction flow limit determining unit 280 may include: and a transaction flow limit judging module. The transaction current limiting judgment module is used for: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the performance data table before updating exceeds the set hot spot threshold value, if so, determining that the current limitation needs to be performed on the transaction corresponding to the transaction request packet.

Fig. 14 is a schematic structural diagram of a monitoring and processing device for hot spot records in a real-time processing system according to yet another embodiment of the present invention. As shown in fig. 14, the monitoring and processing device for hot spot recording in the real-time processing system shown in fig. 13 may further include: the performance data search unit 2100 and the performance data validity determination unit 2110 are connected to each other, both connected to the request packet forwarding unit 290, and respectively connected to the request packet interception unit 210 and the transaction flow limit determination unit 280.

The performance data search unit 2100 is configured to: and judging whether a performance data record related to the record information exists in the performance data by searching the performance data table.

The performance data validity judging unit 2110 is configured to: if not, the transaction request packet is forwarded to the real-time processing system for logic processing; if yes, judging whether the performance data record of the record information meets the following conditions: and if not, forwarding the transaction request packet to the real-time processing system for logic processing, and if so, judging whether the flow limitation needs to be performed on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot point threshold value.

In this embodiment, if the performance data table determines that there is no performance data record related to the record information in the transaction request packet, the transaction request packet may be directly forwarded to the real-time processing system for logic processing. If the performance data table determines that there is a performance data record related to the record information in the transaction request packet, it may continue to check whether the performance data record is valid, and the determination method may be: the end time of the statistical window is greater than the current time point-the duration of the statistical window and the number of the statistical windows. The data of the statistic window ending time field in the performance data table may be the statistic window ending time stored in the current performance data table. The current time point may be a time point of the time, for example, a time point of the judgment time point may be a time difference between the time point and 1 month 1 day 00:00:00 of 1900 year, and the unit may be milliseconds. If the performance data record is judged to be invalid, which indicates that the current time point is not in the statistical window, the transaction request packet can be directly forwarded to the real-time processing system for logic processing, and the transaction is not counted in the statistical range, so that the transaction in the statistical window can be guaranteed to be counted. If the performance data record is judged to be valid, a hot spot threshold value (current limiting condition) is set to judge whether to limit the current.

Fig. 15 is a schematic structural diagram of a monitoring and processing device for hot spot records in a real-time processing system according to still another embodiment of the present invention. As shown in fig. 15, the monitoring and processing device for hot spot recording in the real-time processing system shown in fig. 13 may further include: the transaction block determining unit 2120 and the transaction block processing unit 2130 are connected to each other and are respectively connected to the transaction flow limit determining unit 280 and the request packet forwarding unit 290, and the transaction block processing unit 2130 is further connected to the performance data updating unit 250.

The transaction block determining unit 2120 is configured to: if the transaction corresponding to the transaction request packet needs to be limited, judging whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in a current limiting window corresponding to the record information in the performance data table before updating.

The transaction blocking processing unit 2130 is configured to: if yes, generating a transaction rejection packet according to the transaction request packet, taking the time point generated by the transaction rejection packet as a return timestamp, and updating a performance data table according to the record information in the transaction request packet and the return timestamp; if not, updating the data of the forwarded transaction number column in the current limiting window corresponding to the record information in the performance data table, and executing to forward the transaction request packet to the real-time processing system for logic processing.

In the transaction blocking unit 2130, the transaction that needs to be throttled may be generated into a return packet of a special format, and the reject packet may be a special form of the return packet, and may also notify the receiving party that the system is busy and the request is not accepted. If the real-time processing system needs to be prevented from processing the transaction request packet, after a transaction rejection packet is generated, updating the performance data table according to the record information in the transaction request packet and the return timestamp can be executed, the transaction corresponding to the transaction rejection packet can be regarded as an abnormal transaction, and accordingly data of an abnormal transaction number field in a statistical window in the performance data table can be updated, so that the transaction rejection packet can also be used for hotspot analysis and positioning, and the analysis result is more accurate. When the performance data updating unit 250 acts, the return timestamp can be replaced by the time point when the transaction rejection packet is generated, and the record information in the transaction request packet can be obtained from the transaction rejection packet or the information temporarily stored when the transaction request packet is accepted. The reject packet may be returned to the location in the online process where the information is needed. If the transaction request packet does not need to be prevented from being processed by the real-time processing system, the data of the forwarded transaction number field in the current limiting window corresponding to the record information in the performance data table can be updated, for example, one is added, and the transaction request packet is forwarded to the real-time processing system for logic processing.

In some embodiments, the setting the current limit parameter comprises: the time length of the current-limiting window, the number of the current-limiting windows and the number of transactions allowed to be forwarded by each current-limiting window; the performance data table further includes a number of forwarded transaction number fields in the current limiting windows, where the number of forwarded transaction number fields is the number of the current limiting windows, and the current limiting windows of the forwarded transaction number fields in each current limiting window have the duration of the current limiting window and are arranged in time sequence. The transaction block determining unit 2120 may include: a transaction block determination module. A transaction block determination module to: and judging whether the real-time processing system needs to be prevented from processing the transaction request packet or not according to whether the data of the forwarded transaction number field in each current limiting window corresponding to the record information in the performance data table before updating is larger than or equal to the number of the transactions allowed to be forwarded by the current limiting window.

Fig. 16 is a schematic structural diagram of a monitoring and processing device for hot spot records in a real-time processing system according to still another embodiment of the present invention. As shown in fig. 16, the monitoring and processing device for hot spot recording in the real-time processing system shown in fig. 15 may further include: the current limit window ending time updating unit 2140 is connected between the transaction current limit determining unit 280 and the transaction block determining unit 2120.

The current limit window ending time updating unit 2140 is configured to: and judging whether the current time point is less than or equal to the data of the current limiting window ending time column corresponding to the record information in the performance data table, if not, updating the data of the current limiting window ending time column according to the current time point and the set current limiting window duration.

In some embodiments, the performance data updating unit 250 may further be configured to: the performance data table comprises a plurality of abnormal transaction quantity columns in the statistical window, and the statistical windows of the abnormal transaction quantity columns in the statistical windows are arranged in a time sequence. In this embodiment, the time may be divided into a plurality of continuous windows according to the value specified by the "statistical window duration" parameter, so that the statistical data in the performance data table may better conform to the actual processing condition of the real-time processing system, and the accuracy of hotspot positioning and current limiting may be further improved.

In some embodiments, the hot spot record determining unit 260 may include: and a hotspot record judging module. The hotspot record judging module is used for: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the updated performance data table exceeds the set hot spot threshold value, and if so, confirming that a hot spot appears on the transaction type and the record information.

Fig. 17 is a schematic structural diagram of a performance data updating unit according to an embodiment of the present invention. As shown in fig. 17, the performance data updating unit 250 may include: a record-related field acquisition module 251, a first field data update module 252, and a second field data update module 253, which are connected in sequence.

The registration-related field acquiring module 251 is configured to: and acquiring a field of the end time of the statistical window and a field of the number of abnormal transactions in the statistical window related to the record information in the performance data table.

The first field data updating module 252 is configured to: and calculating new statistical window ending time according to the return timestamp and the set statistical window duration so as to enable the return timestamp to be smaller than the new statistical window ending time, and updating the acquired data of the statistical window ending time column by utilizing the new statistical window ending time.

The second field data updating module 253 is configured to: calculating the new end time of the statistical window of the abnormal transaction quantity column in each acquired statistical window according to the new end time of the statistical window and the set statistical window duration; comparing whether the new end time is the same as the window end time corresponding to the abnormal transaction number column in the obtained statistical window or not; if the two end times are the same, updating the data of the abnormal transaction number field in the obtained statistical window corresponding to the window end time by using the abnormal transaction number field in the obtained statistical window corresponding to the new end time; and adding one to the data of the abnormal transaction number field in the statistical window acquired at the last time, and updating the data of the abnormal transaction number field in the acquired statistical window which is not updated to zero.

In this embodiment, the first field data update module 252 may continuously update the end time of the statistical window (the end time of the total statistical period/the end time of the latest statistical window), and the second field data update module 253 may further continuously update the end time of each statistical window. The second field data update module 253 can further update the data of the abnormal transaction number field in the statistical window in the performance data table according to the updated end time (new end time) of each statistical window. And when the transaction of the transaction request packet processing is considered to belong to the abnormal transaction, adding one to the data of the abnormal transaction number field in the statistical window acquired at the last time. Therefore, the statistical data in the performance data table can be ensured to be the data which is the nearest to the processing time of the request packet, so that the transaction processing performance condition of the real-time processing system can be reflected better in real time.

Fig. 18 is a schematic structural diagram of a registration-related field acquisition module according to an embodiment of the present invention. As shown in fig. 18, the registration-related field acquiring module 251 may include: a hotspot record field search module 2511 and a record related field acquisition module 2512, which are connected to each other.

The hotspot record field search module 2511 is configured to: and judging whether the performance data table contains a hotspot record information column corresponding to the record information or not by searching the performance data table.

The record-related field acquisition module 2512 is configured to: if yes, obtaining a statistical window ending time field and an abnormal transaction quantity field in the statistical window related to the record information according to the searched hotspot record information field.

In this embodiment, for example, the performance data table may be searched using information such as configuration codes and record information of transactions, if records of the same contents such as configuration codes and record information are searched in the performance data table, the search may be completed, and a statistical window end time field and an abnormal transaction number field in the statistical window related to the record information may be obtained according to the searched hotspot record information field. And obtaining the field related to the record information in the performance data table by searching, so that the data of the related field in the performance data table can be directly updated.

Fig. 19 is a schematic structural diagram of a registration-related field acquisition module according to another embodiment of the invention. As shown in fig. 19, the registration-related field retrieving module 251 shown in fig. 18 may further include: a record-related field adding module 2513 connected to the hot spot record field searching module 2511.

The record associated field addition module 2513 is configured to: if the performance data table does not contain the hotspot record information column corresponding to the record information, adding the hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window related to the record information in the performance data table, and initializing the added hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window.

In this embodiment, the added data of the hotspot record information field may be the record information, and the field of the end time of the statistical window and the field of the number of abnormal transactions in the statistical window may be initialized to zero, for example. By adding a piece of performance data record in the performance data table, performance record data can be added for the transaction which is not recorded in the performance data table before.

Fig. 20 is a schematic structural diagram of a registration-related field acquisition module according to yet another embodiment of the invention. As shown in fig. 20, the record-related field retrieving module 251 shown in fig. 18 may further include: the failure record recycling module 2514 is connected to the hot spot record field searching module 2511.

The failure record recovery module 2514 is configured to: if the performance data table does not contain the hotspot record information column corresponding to the record information, searching and finding a statistic window failure record in the performance data table, initializing the statistic window failure record, and respectively taking the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the initialized statistic window failure record as the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the record information, wherein the statistic window failure record satisfies the following conditions: and returning a timestamp > the end time of the statistical window in the statistical window failure record + the number of the set statistical windows and the set statistical window duration.

In this embodiment, the record in the performance data table satisfying the following condition may be a statistical window failure record: and returning a time point > the end time of the statistical window + the number of the statistical windows x the duration of the statistical windows, if the performance data table does not have the performance data record of the record information, searching the performance data table to find out the record of failure of the first statistical window, wherein the record in the performance data table can be recovered and reused, so that the waste of the storage space of the system can be avoided, and the storage pressure of the system is reduced.

Fig. 21 is a schematic structural diagram of a monitoring and processing device for hot spot records in a real-time processing system according to another embodiment of the present invention. As shown in fig. 21, the monitoring and processing device for hotspot records in the real-time processing system of this embodiment may include a configuration data module (1), a performance data module (2), a data acquisition module (3), an information asynchronous transmission module (4), an analysis module (5), and a hotspot current limiting module (6).

Configuration data module (1): the function is the maintenance of the configuration table contents. The method can include configuring collection conditions, collecting contents, hot spot judgment conditions and current limiting parameters, and can be used for completing the functions of step (a1), step (a2), step (a3), step (a4) and the like shown in fig. 10. The content of the configuration table can be used by the data acquisition module (3), the analysis module (5) and the hot spot current limiting module (6). The configuration table data structure maintained by the configuration data module (1) may be as shown in table 8.

Column position	Type (B)	Remarks for note
			Configuration coding	Character string	Main key
Transaction type	Character string	Main key
			Threshold processing time (ms)	Number of
Input packet start bit 1	Number of
			Input packet end bit 1	Number of
……	……	……
			Input packet start bit n	Number of
End bit n of input packet	Number of
			Statistical window duration (ms)	Number of
Counting the number of windows	Number of
			Hot spot threshold	Number of
Limiting window duration	Number of
			Limiting number of transactions	Number of

Table 8 configuration table data structure

Configuring codes: uniquely identifying a piece of configuration information, configuring key values of a table, and not repeating.

The transaction type: a uniquely identified transaction type. The transaction type encoding value is consistent with the real-time processing system, and the transaction type encoding field should be present in the transaction request packet.

Threshold processing time: the threshold processing time is set according to the normal processing time of various transactions of the real-time processing system and is in unit of millisecond. For normally returned transactions (another portion of transactions are rejected by the current limit module), a processing time greater than the threshold processing time is a necessary condition to be collected.

Input packet start bit, input packet end bit: pairing uses, specifying the location of information that needs to be collected from the incoming packet if screening is satisfied, starting with 1. Because the specific record key values may be scattered in different places in the uplink packet, a plurality of columns can be intercepted for joint acquisition, a plurality of pairs can be configured according to the complex situation of the real-time system, and if the value is 0, the pairing can be indicated to be unused.

Counting the window duration: the size of each time window is set to record how many transaction performance anomalies exist in the time window. In milliseconds.

Counting the number of windows: and setting the number of time windows and recording only the latest information in the time windows. In some embodiments, the time window is sliced from 1900/1/00: 00.000 by dividing the time into a number of consecutive windows according to the value specified by the "statistical window duration" parameter. When the analysis module (5) works, transaction data in a time window of the latest 'number of statistical windows' is always counted and analyzed along with time change, so that the analyzed data are ensured to be in a quasi-real-time state. The statistical windows are numbered from near to past, 1, 2, 3, … …. FIG. 22 is a diagram illustrating a time slicing into multiple consecutive statistics windows according to the duration of the statistics windows in accordance with an embodiment of the present invention. As shown in fig. 22, the start time of the statistical window may be 08:00:15, the end time of the statistical window may be 08:01:15, the time period between the start time of the statistical window and the end time of the statistical window is divided into 12 statistical windows according to the statistical window duration 5s, and the statistical windows are numbered as 1, 2, 3, … …, and 12 in sequence from the nearest to the current time point.

Hot spot threshold: when the sum of the transaction processing performance abnormality and the number of the limited transactions in all the monitoring windows is larger than the threshold value, which can indicate that a hot spot occurs, measures, such as limiting the current, need to be taken.

Duration of current limit window, number of current limit transactions: and setting the number of the allowed current-limiting transactions of the hotspot current-limiting module in the current-limiting window, rejecting the subsequent transactions with the number exceeding the number of the current-limiting transactions, and not forwarding the subsequent transactions to the real-time processing system. The time may be divided into a plurality of consecutive windows according to the value specified by the "time duration of the current limiting window" parameter, starting at 1900 s 1/00: 00: 00.000. When the current limit window duration is 0, it may indicate that current limit is not performed.

Performance data module (2): a table of performance data is stored. The content of the performance data table is maintained and used by the analysis module (5) and the hot spot current limiting module (6). The data structure of the performance data table may be as shown in table 9.

Column position	Type (B)	Remarks for note
			Configuration coding	Character string	Main key
Hotspot record information	Character string	Main key
			Statistical Window
1 end time	Number of
			Counting the number of abnormal transactions in window 1	Number of
Count number of anomalous transactions within Window 2	Number of
			……	……	……
Counting the number of abnormal transactions in the window n	Number of
			End time of current limit window	Number of
Forwarded transaction number within a current limit window	Number of

Table 9 structure of performance data table

Configuring codes: the key value of a configuration table can be uniquely identified, and the configuration information and the key value cannot be repeated.

And recording information in the hot spot. The content in the transaction request packet can be intercepted according to the start bit and the end bit of the input packet in the configuration table, and when a plurality of pairs of information of the start bit and the end bit of the input packet exist in the configuration table, the plurality of pairs can be spliced according to the sequence.

End time of statistical window 1: the time may be divided into a plurality of consecutive windows according to the value specified by the "statistical window duration" parameter starting from 1900/1/00: 00: 00.000. The end time of the window of return packet time points for which an exception transaction has recently occurred may be included. If the return packet time point of the most recent exception transaction is x, there may be:

statistical window 1 end time (round (x/statistical window duration) +1) statistical window duration

Counting the number of abnormal transactions in the window: the statistical windows may be numbered sequentially from the near past, 1, 2, 3, … …. The transaction return packet time point is the sum of the number of performance exception transactions and the number of transactions that are throttled within each "statistical window".

End time of current limit window: the time may be divided into a plurality of consecutive windows according to the value specified by the "time duration of the current limiting window" parameter, starting at 1900 s 1/00: 00: 00.000. The field may point to the end time of the current limit window, and as time passes, the current limit module will adjust the end time to point to the nearest current limit window.

Number of current limiting transactions: the number of transaction requests allowed to pass through each current limit window. The exceeding transactions will be directly rejected by the flow restriction module to relieve the real-time processing system of stress.

Data acquisition module (3): the system can be responsible for intercepting and forwarding a transaction request packet, a transaction return packet and a transaction reject packet, judging whether information needs to be acquired or not according to configuration table information, acquiring the information needing to be acquired and sending the information to a subsequent module in an asynchronous mode. Can be used for completing the functions of step (b1), (b2), (b6), (b7), (b8) and (b9) in the method shown in FIG. 10.

The module inputs may include: configuration tables, transaction send-up packets, transaction return packets, transaction reject packets, and the like.

The module output may include: a transaction return packet, a transaction reject packet, information collection, etc.

The data structure of the acquisition information may be as shown in table 10.

Column position	Type (B)	Remarks for note
			Configuration coding	Character string	Main key
Hotspot record information	Character string	Main key
			Point of time of return	Number of

Table 10 data structure for collecting information

Hotspot recording information: the content in the transaction request packet can be intercepted according to the start bit and the end bit of the input packet in the configuration table, and when a plurality of pairs of information of the start bit and the end bit of the input packet exist in the configuration table, the plurality of pairs are spliced according to the character strings.

Returning to a time point: the time point of transaction return is at intervals of 00:00:00.000 units of milliseconds from 1 month 1 of 1900 years.

Information asynchronous transmission module (4): may be responsible for asynchronous transfer of information and may be implemented using sophisticated asynchronous transfer middleware such as MQ.

The input and output of the module are collected information.

Analysis module (5): the current limiting module can be responsible for receiving information, performing statistical analysis, and maintaining the result in a performance data table for current limiting judgment reference of the current limiting module. The performance data table may be loaded into memory to improve efficiency. The analysis module can be operated in a resident mode or triggered to operate by the information asynchronous transmission module. The function of step (b10) in the method description is completed.

The module inputs may include: a configuration table; and collecting information.

The module output may include: and (5) a performance data table.

Hot spot current limiting module (6): the system can be responsible for receiving the transaction request packet forwarded by the data acquisition module, determining whether to carry out hot point current limiting or remove the hot point current limiting according to the configuration table and the performance data table, and updating the current limiting data in the performance data table. Can be used for completing the functions of the step (b3), the step (b4) and the step (b5) in the method shown in FIG. 10.

The module inputs may include: a transaction uploading packet, a transaction returning packet, a configuration table and a performance data table.

The module output may include: a transaction uploading packet, a transaction returning packet, a transaction rejecting packet and a performance data table.

Fig. 23 is a schematic diagram of a monitoring and processing device for hotspot recording deployed outside a real-time processing system in a real-time processing system according to an embodiment of the invention. FIG. 24 is a schematic diagram of a real-time processing system with a hotspot record monitoring and processing device deployed therein according to an embodiment of the invention. As shown in fig. 23 and 24, the apparatus according to the embodiment of the present invention may be deployed in two ways, one is external to the real-time processing system, and the other is internal to the real-time processing system. The two deployment modes have the same monitoring effect, the flexibility and the portability of the external deployment mode are strong, but the performance is slightly influenced due to the time consumption of transmission between the device and the real-time system. The built-in deployment mode has better performance, but has slightly poor monitoring flexibility and portability.

The embodiment of the invention provides a method for flexibly acquiring hot spot record specific information from a transaction request packet and a transaction return packet in a configuration mode. The hot spot record related information comprises transaction processing time, transaction type and specific information capable of identifying records. By utilizing the correlation between the hotspot records and the slow collective processing, the specific hotspot records can be found by a statistical method. Hotspot records can cause the collective performance of related transactions on the record to degrade; sudden appearance of a certain transaction on a certain record (such as 60 seconds) and the integration performance decline can basically exclude other occasional interference factors, and the temporary hot spot phenomenon exists on the records. The asynchronous processing mode of the analysis module can ensure that the real-time processing system has little influence on the transaction processing. The method of slicing time is adopted to form continuous time windows, the transaction condition is monitored by a statistical method of a plurality of recent continuous time windows, and the analyzed data is guaranteed to continuously reflect the latest state in real time. The current limiting is specific to specific transaction types and specific hot spot records, and has good accuracy. The current limiting method adopts a total method in a time window, and can effectively utilize the buffering capacity of the real-time processing system to transactions under the condition of not causing congestion of the real-time processing system, so as to process the transactions as much as possible. The method can complete monitoring by installing a device at the inlet of a real-time system, has very little influence on the existing system and has strong applicability. The whole process can be automated, real-time and configurable. The device can accomplish the monitoring at a device of real-time system entrance installation, and is very little to current system's influence, and the suitability is strong. The device is only transformed at a front point of the real-time system, the existing real-time system does not need to be changed at all, and the cost is low. Hot spot records appearing in the real-time processing system can be quickly and accurately positioned; automatically and accurately limiting the current of the request transaction causing the hotspot record, and cooling the hotspot; and when the transaction request for operating the hotspot record is reduced, the current limiting measure is automatically removed, so that the safe and stable operation of the system is ensured. The system has the characteristics of quick response, accurate positioning, automatic disposal, convenient modification, strong universality and high performance, plays an important role in ensuring the smooth operation of large-scale important real-time processing systems (such as banks, securities and public utilities) and preventing the system from being blocked or even collapsing, and has good social benefit and economic benefit.

According to the monitoring and processing device for the hot spot records in the real-time processing system, the performance data table is updated according to the record information and the return timestamp in the transaction request packet, whether hot spots occur on the transaction type and the record information corresponding to the transaction request packet is judged according to the data of the abnormal transaction number field in the statistical window in the updated performance data table and the set hot spot threshold value, the hot spots caused by frequent operation of which type of transactions on which specific record can be accurately positioned, the positioning precision is high, and other unrelated services or unrelated records cannot be influenced by treatment based on the hot spots. The hotspot analysis and positioning can be performed by receiving the transaction request packet and the transaction return packet and acquiring corresponding information, which is irrelevant to the internal architecture of a specific real-time processing system, so that the hotspot analysis and positioning method is generally applicable to various real-time processing systems and has strong universality. The hotspot analysis of the embodiment is specific to each transaction, when the hotspot record appears, the positioning can be rapidly completed, and the reaction speed can reach the second level by setting appropriate parameters, so that the reaction is rapidly reflected.

The method and the device of the embodiment of the invention can quickly and accurately position the hot spot record in the real-time processing system; automatically and accurately limiting the current of the request transaction causing the hotspot record, and cooling the hotspot; and when the transaction request for operating the hotspot record is reduced, the current limiting measure is automatically removed, so that the safe and stable operation of the system is ensured. The system has the characteristics of quick response, accurate positioning, automatic disposal, convenient modification, strong universality and high performance, plays an important role in ensuring the smooth operation of large-scale important real-time processing systems (such as banks, securities and public utilities) and preventing the system from being blocked or even collapsing, and has good social benefit and economic benefit.

In the description herein, reference to the description of the terms "one embodiment," "a particular embodiment," "some embodiments," "for example," "an example," "a particular example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. The sequence of steps involved in the various embodiments is provided to schematically illustrate the practice of the invention, and the sequence of steps is not limited and can be suitably adjusted as desired.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are only exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims

1. A method for monitoring and processing hotspot records in a real-time processing system, comprising:

receiving a transaction request packet and recording a receiving time stamp of the transaction request packet;

receiving a transaction return packet returned after the real-time processing system processes the transaction request packet, and recording a return timestamp of the transaction return packet;

calculating the transaction response time according to the receiving timestamp and the returning timestamp;

judging whether the transaction response time exceeds a set threshold processing time or not;

if so, updating a performance data table according to the record information in the transaction request packet and the return timestamp, wherein the performance data table comprises a hotspot record information column corresponding to the record information, a statistic window ending time column and an abnormal transaction number column in a statistic window;

judging whether a hot spot appears on the transaction type corresponding to the transaction request packet and the record information according to the updated data of the abnormal transaction number column in the statistical window in the performance data table and a set hot spot threshold value;

if yes, limiting the transaction types and the transaction of the record information according to set current limiting parameters;

before receiving a transaction return packet returned after the real-time processing system processes the transaction request packet, the method further includes:

judging whether the flow limitation needs to be carried out on the transaction corresponding to the transaction request packet or not according to the data of the abnormal transaction number column in the statistical window in the performance data table before updating and the set hot spot threshold value;

if not, forwarding the transaction request packet to the real-time processing system for logic processing;

before judging whether the current limitation needs to be performed on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot spot threshold, the method further includes:

judging whether a performance data record related to the record information exists in the performance data by searching the performance data table;

if not, the step of forwarding the transaction request packet to the real-time processing system for logic processing is executed; if yes, judging whether the performance data record of the record information meets the following conditions: and if not, forwarding the transaction request packet to the real-time processing system for logic processing, and if so, judging whether the flow limitation needs to be performed on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot point threshold value.

2. The method for monitoring and processing hot spot records in a real-time processing system as claimed in claim 1, further comprising:

if the transaction corresponding to the transaction request packet needs to be limited, judging whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in a current limiting window corresponding to the record information in the performance data table before updating;

if yes, generating a transaction rejection packet according to the transaction request packet, taking the time point generated by the transaction rejection packet as a return timestamp, and executing the step of updating a performance data table according to the record information in the transaction request packet and the return timestamp; if not, updating the data of the forwarded transaction number column in the current limiting window corresponding to the record information in the performance data table, and executing the step of forwarding the transaction request packet to the real-time processing system for logic processing.

3. The method for monitoring and processing hot spot records in a real-time processing system according to claim 2, wherein before determining whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in the current limiting window corresponding to the record information in the performance data table before updating, the method further comprises:

and judging whether the current time point is less than or equal to the data of the current limiting window ending time column corresponding to the record information in the performance data table, if not, updating the data of the current limiting window ending time column according to the current time point and the set current limiting window duration.

4. The method as claimed in claim 1, wherein the performance data table comprises a plurality of abnormal transaction number fields in the statistical window, and the statistical window of the abnormal transaction number field in each statistical window is arranged in time sequence.

5. The method of claim 4, wherein updating the performance data table according to the record information in the transaction request packet and the return timestamp comprises:

acquiring a statistical window ending time column and an abnormal transaction quantity column in a statistical window related to the record information in the performance data table;

calculating new statistical window ending time according to the return timestamp and the set statistical window duration so that the return timestamp is smaller than the new statistical window ending time, and updating the data of the acquired statistical window ending time column by using the new statistical window ending time;

calculating the new end time of the statistical window of the abnormal transaction quantity column in each acquired statistical window according to the new end time of the statistical window and the set statistical window duration; comparing whether the new end time is the same as the window end time corresponding to the abnormal transaction number column in the obtained statistical window or not; if the two end times are the same, updating the data of the abnormal transaction number field in the obtained statistical window corresponding to the window end time by using the abnormal transaction number field in the obtained statistical window corresponding to the new end time; and adding one to the data of the abnormal transaction number field in the statistical window acquired at the last time, and updating the data of the abnormal transaction number field in the acquired statistical window which is not updated to zero.

6. The method of claim 5, wherein the obtaining of the end time field of the statistical window and the number of abnormal transactions in the statistical window of the performance data table with respect to the record information comprises:

judging whether the performance data table contains a hotspot recording information column corresponding to the recording information or not by searching the performance data table;

if yes, obtaining a statistical window ending time field and an abnormal transaction quantity field in the statistical window related to the record information according to the searched hotspot record information field.

7. The method for monitoring and processing hot spot records in a real-time processing system as claimed in claim 6, wherein the obtaining of the end time field of the statistical window and the number of abnormal transactions in the statistical window about the record information in the performance data table further comprises:

if the performance data table does not contain the hotspot record information column corresponding to the record information, adding the hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window related to the record information in the performance data table, and initializing the added hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window.

8. The method for monitoring and processing hot spot records in a real-time processing system as claimed in claim 6, wherein the obtaining of the end time field of the statistical window and the number of abnormal transactions in the statistical window about the record information in the performance data table further comprises:

if the performance data table does not contain the hotspot record information column corresponding to the record information, searching and finding a statistic window failure record in the performance data table, initializing the statistic window failure record, and respectively taking the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the initialized statistic window failure record as the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the record information, wherein the statistic window failure record satisfies the following conditions: and returning a timestamp > the end time of the statistical window in the statistical window failure record + the number of the set statistical windows and the set statistical window duration.

9. The method for monitoring and processing hot spot records in a real-time processing system according to claim 4, wherein judging whether a hot spot occurs in the transaction type and the record information corresponding to the transaction request packet according to the updated data of the abnormal transaction number field in the statistical window in the performance data table and a set hot spot threshold comprises:

and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the updated performance data table exceeds the set hot spot threshold value, and if so, confirming that a hot spot appears on the transaction type and the record information.

10. The method of claim 1, wherein the performance data table comprises a plurality of abnormal transaction number fields in the statistical window, and the statistical window of the abnormal transaction number field in each statistical window is arranged in time sequence; judging whether the flow limitation needs to be performed on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot spot threshold value, wherein the judging step comprises the following steps:

and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the performance data table before updating exceeds the set hot spot threshold value, if so, determining that the current limitation needs to be performed on the transaction corresponding to the transaction request packet.

11. The method for monitoring and processing hot spot records in a real-time processing system as claimed in claim 2, wherein said setting the current limiting parameter comprises: the time length of the current-limiting window, the number of the current-limiting windows and the number of transactions allowed to be forwarded by each current-limiting window; the performance data table further comprises a forwarded transaction number field in a plurality of current limiting windows, the number of the forwarded transaction number fields being the number of the current limiting windows, and the current limiting windows of the forwarded transaction number fields in each current limiting window have the duration of the current limiting window and are arranged in time sequence; judging whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in the current limiting window corresponding to the record information in the performance data table before updating, wherein the judging comprises the following steps:

and judging whether the real-time processing system needs to be prevented from processing the transaction request packet or not according to whether the data of the forwarded transaction number field in each current limiting window corresponding to the record information in the performance data table before updating is larger than or equal to the number of the transactions allowed to be forwarded by the current limiting window.

12. The method of claim 1, wherein the transaction type is represented by data of a transaction type encoding field, and the transaction type encoding field is pre-added to the transaction request packet.

13. The method for monitoring and processing hot spot records in a real-time processing system as claimed in claim 1, wherein the record information is obtained based on the start bit and the end bit of the input packet.

14. A device for monitoring and processing hotspot records in a real-time processing system, comprising:

a request packet interception unit for: receiving a transaction request packet and recording a receiving time stamp of the transaction request packet;

a return packet interception unit for: receiving a transaction return packet returned after the real-time processing system processes the transaction request packet, and recording a return timestamp of the transaction return packet;

a response time calculation unit for: calculating the transaction response time according to the receiving timestamp and the returning timestamp;

a collected information judging unit configured to: judging whether the transaction response time exceeds a set threshold processing time or not;

a performance data update unit to: if so, updating a performance data table according to the record information in the transaction request packet and the return timestamp, wherein the performance data table comprises a hotspot record information column corresponding to the record information, a statistic window ending time column and an abnormal transaction number column in a statistic window;

a hotspot record judging unit, configured to: judging whether a hot spot appears on the transaction type corresponding to the transaction request packet and the record information according to the updated data of the abnormal transaction number column in the statistical window in the performance data table and a set hot spot threshold value;

a hotspot record processing unit, configured to: if yes, limiting the transaction types and the transaction of the record information according to set current limiting parameters;

a transaction flow limit determination unit configured to: judging whether the flow limitation needs to be carried out on the transaction corresponding to the transaction request packet or not according to the data of the abnormal transaction number column in the statistical window in the performance data table before updating and the set hot spot threshold value;

a request packet forwarding unit configured to: if not, forwarding the transaction request packet to the real-time processing system for logic processing;

a performance data search unit to: judging whether a performance data record related to the record information exists in the performance data by searching the performance data table;

a performance data validity judgment unit configured to: if not, the transaction request packet is forwarded to the real-time processing system for logic processing; if yes, judging whether the performance data record of the record information meets the following conditions: and if not, forwarding the transaction request packet to the real-time processing system for logic processing, and if so, judging whether the flow limitation needs to be performed on the transaction corresponding to the transaction request packet according to the data of the abnormal transaction number field in the statistical window in the performance data table before updating and the set hot point threshold value.

15. The apparatus for monitoring and processing hot spot records in a real-time processing system as claimed in claim 14, further comprising:

a transaction block determination unit to: if the transaction corresponding to the transaction request packet needs to be limited, judging whether the real-time processing system needs to be prevented from processing the transaction request packet according to the set current limiting parameter and the number of forwarded transactions in a current limiting window corresponding to the record information in the performance data table before updating;

a transaction blocking processing unit to: if yes, generating a transaction rejection packet according to the transaction request packet, taking the time point generated by the transaction rejection packet as a return timestamp, and updating a performance data table according to the record information in the transaction request packet and the return timestamp; if not, updating the data of the forwarded transaction number column in the current limiting window corresponding to the record information in the performance data table, and executing to forward the transaction request packet to the real-time processing system for logic processing.

16. The apparatus for monitoring and processing hot spot records in a real-time processing system as claimed in claim 15, further comprising:

a current limit window end time update unit for: and judging whether the current time point is less than or equal to the data of the current limiting window ending time column corresponding to the record information in the performance data table, if not, updating the data of the current limiting window ending time column according to the current time point and the set current limiting window duration.

17. The apparatus for monitoring and processing hot spot records in a real-time processing system as claimed in claim 14, wherein said performance data updating unit is further configured to: the performance data table comprises a plurality of abnormal transaction quantity columns in the statistical window, and the statistical windows of the abnormal transaction quantity columns in the statistical windows are arranged in a time sequence.

18. The apparatus for monitoring and processing hot spot records in a real-time processing system as claimed in claim 17, wherein said performance data updating unit comprises:

a registration-related field acquisition module configured to: acquiring a statistical window ending time column and an abnormal transaction quantity column in a statistical window related to the record information in the performance data table;

a first field data update module configured to: calculating new statistical window ending time according to the return timestamp and the set statistical window duration so that the return timestamp is smaller than the new statistical window ending time, and updating the data of the acquired statistical window ending time column by using the new statistical window ending time;

a second field data update module configured to: calculating the new end time of the statistical window of the abnormal transaction quantity column in each acquired statistical window according to the new end time of the statistical window and the set statistical window duration; comparing whether the new end time is the same as the window end time corresponding to the abnormal transaction number column in the obtained statistical window or not; if the two end times are the same, updating the data of the abnormal transaction number field in the obtained statistical window corresponding to the window end time by using the abnormal transaction number field in the obtained statistical window corresponding to the new end time; and adding one to the data of the abnormal transaction number field in the statistical window acquired at the last time, and updating the data of the abnormal transaction number field in the acquired statistical window which is not updated to zero.

19. The apparatus for monitoring and processing hot spot logs in a real-time processing system as claimed in claim 18, wherein said log-related field retrieving module comprises:

the hotspot record column searching module is used for: judging whether the performance data table contains a hotspot recording information column corresponding to the recording information or not by searching the performance data table;

a registration-related field acquisition module configured to: if yes, obtaining a statistical window ending time field and an abnormal transaction quantity field in the statistical window related to the record information according to the searched hotspot record information field.

20. The apparatus for monitoring and processing hot spot logs in a real-time processing system as claimed in claim 19, wherein said log-related field retrieving module further comprises:

a record-related field addition module configured to: if the performance data table does not contain the hotspot record information column corresponding to the record information, adding the hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window related to the record information in the performance data table, and initializing the added hotspot record information column, the statistics window ending time column and the abnormal transaction number column in the statistics window.

21. The apparatus for monitoring and processing hot spot logs in a real-time processing system as claimed in claim 19, wherein said log-related field retrieving module further comprises:

a failure record recovery module to: if the performance data table does not contain the hotspot record information column corresponding to the record information, searching and finding a statistic window failure record in the performance data table, initializing the statistic window failure record, and respectively taking the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the initialized statistic window failure record as the hotspot record information column, the statistic window ending time column and the abnormal transaction number column in the statistic window of the record information, wherein the statistic window failure record satisfies the following conditions: and returning a timestamp > the end time of the statistical window in the statistical window failure record + the number of the set statistical windows and the set statistical window duration.

22. The apparatus for monitoring and processing hot spot records in a real-time processing system according to claim 17, wherein the hot spot record determining unit comprises:

the hotspot record judging module is used for: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the updated performance data table exceeds the set hot spot threshold value, and if so, confirming that a hot spot appears on the transaction type and the record information.

23. The apparatus for monitoring and processing hot spot records in a real-time processing system as claimed in claim 14, wherein said performance data table comprises a plurality of abnormal transaction number fields in said statistical window, the statistical window of the abnormal transaction number field in each of said statistical windows being arranged in time sequence; the transaction flow limit judging unit comprises:

a transaction flow limit determination module to: and judging whether the sum of the data of the abnormal transaction number column in each statistical window in the performance data table before updating exceeds the set hot spot threshold value, if so, determining that the current limitation needs to be performed on the transaction corresponding to the transaction request packet.

24. The apparatus for monitoring and processing hot spot records in a real-time processing system as claimed in claim 15, wherein said setting the current limiting parameter comprises: the time length of the current-limiting window, the number of the current-limiting windows and the number of transactions allowed to be forwarded by each current-limiting window; the performance data table further comprises a forwarded transaction number field in a plurality of current limiting windows, the number of the forwarded transaction number fields being the number of the current limiting windows, and the current limiting windows of the forwarded transaction number fields in each current limiting window have the duration of the current limiting window and are arranged in time sequence; the transaction block judgment unit includes:

a transaction block determination module to: and judging whether the real-time processing system needs to be prevented from processing the transaction request packet or not according to whether the data of the forwarded transaction number field in each current limiting window corresponding to the record information in the performance data table before updating is larger than or equal to the number of the transactions allowed to be forwarded by the current limiting window.

25. The apparatus for monitoring and processing hot spot records in a real-time processing system as claimed in claim 14, wherein said performance data updating unit is further configured to: the transaction type is represented by data of a transaction type encoding field, which is pre-added to the transaction request packet.

26. The apparatus for monitoring and processing hot spot records in a real-time processing system as claimed in claim 14, wherein said performance data updating unit is further configured to: and the recording information is obtained according to the start bit and the end bit of the input packet.