Product Features
-
Original transaction analysisPlayback comparisonPlayback configuration
-
- 01
- 02
- 03
- 04
- 05
- 06
Transaction summaryStatistics the transaction summary data of the original data, and you can view the transaction details, specific to field level data.Return code summaryStatistics the return code summary data of the original data. You can query the business module, transaction code and field details corresponding to the return code.ThroughputDisplays the number of transactions, requests and replies per second of the original data.Response timeDisplay the maximum response time, minimum response time and average response time of the original data per second.Top slow transactionQuery the N slowest transactions of the original data.Transaction cloud chartQuery the N busiest transactions of the original data, abnormal return code distribution and slow transaction distribution. -
- 01
- 02
- 03
- 04
- 05
- 06
Transaction code filteringSelect the black-and-white list mode to filter the transaction code, replay the transactions of the transaction code of the white list, and filter the transactions of the transaction code of the blacklist.Time filteringSelect part of the original data according to the time for playback.IP filteringSelect the black-and-white list mode for IP filtering, playback the IP transactions of the white list, and filter the IP transactions of the blacklist.Rate configurationSet the playback speed and maximum TPS upper limit.Sn replacement configurationMark and map the fields to be replaced.Fuse mechanism configurationSet the fuse conditions according to the return code and response time. - 4
-
- 01
- 02
- 03
- 04
- 05
- 06
Transaction coverageQuery the transaction code type and coverage of replayed transactions.Transaction return code comparisonCompare the original response return code and playback response return code of the same request transaction, and count the comparison success number, success rate, failure number, failure rate, etc.Response time deviationQuery the response time deviation and transaction details of all transactions at different time levels.TPS deviationCompare TPS per second of original data with TPS per second of playback data.Top slow analysisQuery the distribution and details of playback transactions whose response time is greater than the threshold.Exception analysisMake statistics on the proportion of wrong transactions and transaction details according to business module, transaction code and return code.
Architecture Deployment

Relevant cases
/ Case introduction
/ Project background
Every year before major promotions (such as "double 11"), banks will evaluate the performance and capacity of databases related to online banking and mobile phones,
and need to conduct stress tests on databases. However, the stress tests of each database are time-consuming and labor-consuming,
resulting in a great waste of manpower and high requirements for personnel skills. Not only that,Each test must enter the ECC machine room for operation,
which not only increases the management and resource consumption of the ECC machine room, but also adds a security risk.
It has become an urgent need to realize the automation of high simulation test and improve the test efficiency. / Customer pain points 1. There are too many high imitation test databases, especially the databases related to "double 11", and it takes 7 days for one person to test one database.
2.The high simulation test has to go to ECC for operation every time, which conflicts with the management of "no or less ECC".
3. The test report needs to be written manually, and many data can be found by themselves. / Solution 1. What are the difficulties in realizing one key high imitation2.1. What are the difficulties in realizing one key high imitation? A high simulation test involves multi center and multi system cooperation, and the test is carried out in the high simulation environment, which is isolated from the production environment;Traffic acquisition and database recovery also need to open a special network.The main difficulties in realizing one key high imitation are as follows:
1) How to conduct cross center and cross network control high simulation test?
2) How to automatically flow from production to quasi production drill environment?
3) How to control the whole process?
4) How to realize the operation control of office network?
5) How to realize large-scale elastic deployment? 2. “"Total score" structure The general control service is deployed in the production environment and opens the access rights of the sub control service of the scheme high simulation environment, so that the general control can control the sub control of packet capture and high simulation test.
3. Database backup and recovery, cross center collaboration2.3. Database backup and recovery, cross center collaboration
The unified backup storage center is adopted, and the production data is backed up to the backup center at a unified fixed point.
The high imitation exercise environment can use the backup files of the center for database high imitation reply,
and the whole process is controlled by the same sub control service and reported to the general control.
4. Automatic flow from production to quasi production drill environment2.4.
Automatic flow from production to quasi production drill environment
Network storage is used for the transmission of traffic packets. The network storage can be accessed in both high imitation environment and production environment.
The production traffic can be automatically transmitted to the network storage,
and the high imitation test can be automatically downloaded to complete the production traffic transmission.
5. High imitation drill and office environment operation2.5. High imitation drill and office environment operation
The high imitation exercise environment is isolated from the office network on the network and cannot be operated directly in the office environment.
The data in the exercise environment is production data, and access is prohibited.
One click high imitation uses the network agent to make the general control web service accessible in the office environment.
The web only provides process control and does not involve sensitive data.
6. Flexible deployment
With more and more high simulation test tasks of the database, it is necessary to complete the test deployment more quickly, conduct the test more simply,
and switch the test tasks more quickly. One click high simulation adopts multi system collaborative operation, which can quickly obtain the production flow required for the test,
control all test processes, and automatically select the playback server and database server;The interface is easy to operate.
Only the database related information needs to be configured. All processes are completed by the background, and the interface provides process status and progress.
/ Program advantages
1. Database test coverage increased by more than 2 times:
Before using one key high imitation, the annual double 11 needs to be tested 4 months in advance, and 10-15 sets of database tests can be completed.
After using one key high imitation, the tests only need to be tested 2 months in advance, and 30-50 sets of database tests can be completed (if resources are sufficient,
all database tests in the bank can be completed in a short time)
2. Save 90% manpower investment:
To realize one click high imitation, it only needs DBA or other high imitation testers to input the relevant test data in the web interface to realize one click test.
The test report can be viewed after the test is completed. The whole process does not need human intervention, reducing the number of special testers.
3. Reduced requirements for human skills:
When one click high imitation is not used, testers need the operation skills of packet capture, FTP, database recovery and high imitation system. After realizing high imitation,
these processes are automated, and these skills are not needed.
4. Annual test database data improvement:
To achieve one click high imitation, as long as there are enough servers in the high imitation environment, you can test as many libraries as you want,
and you don't need to invest extra manpower.
5. 0 access to ECC machine room:
Realize one key high imitation, only need to operate on the office network, and there is no need to enter the ECC machine room for operation.
6. More flexible testing:
Realize one key high imitation, which can be carried out at any time, whether it is pressure test, upgrade or fault recurrence.
It has become an urgent need to realize the automation of high simulation test and improve the test efficiency. / Customer pain points 1. There are too many high imitation test databases, especially the databases related to "double 11", and it takes 7 days for one person to test one database.
2.The high simulation test has to go to ECC for operation every time, which conflicts with the management of "no or less ECC".
3. The test report needs to be written manually, and many data can be found by themselves. / Solution 1. What are the difficulties in realizing one key high imitation2.1. What are the difficulties in realizing one key high imitation? A high simulation test involves multi center and multi system cooperation, and the test is carried out in the high simulation environment, which is isolated from the production environment;Traffic acquisition and database recovery also need to open a special network.The main difficulties in realizing one key high imitation are as follows:
1) How to conduct cross center and cross network control high simulation test?
2) How to automatically flow from production to quasi production drill environment?
3) How to control the whole process?
4) How to realize the operation control of office network?
5) How to realize large-scale elastic deployment? 2. “"Total score" structure The general control service is deployed in the production environment and opens the access rights of the sub control service of the scheme high simulation environment, so that the general control can control the sub control of packet capture and high simulation test.

/ Case introduction
The core business system of a joint-stock bank plans to move down from the original centralized database AS400 / DB2 to x86 distributed database.
Due to the great differences between the old and new core systems, there is an urgent need for a tool that can support rapid simulation regression test to verify
the stability and correctness of the functions of the new core system.
/ Project background
Because the old core system of a joint-stock bank depends on a single system, there are some deficiencies, such as scalability bottleneck, processing capacity bottleneck and so on.
Combined with the needs of business expansion and its overall strategic objectives, the construction of distributed new core system was started in 2017.
Move down from the original as / 400 core system to the x86 core system. In order to ensure the smooth implementation of the core move down,
it is proposed to adopt the simulation parallel process method and use the system simulation tools to realize the parallel operation and maintenance of the old core system
and the new core system, so as to verify the availability and correctness of the new core system and achieve the goal of smooth switching and operation at one time,
Reduce the test investment and production risk of core down moving projects.
/ Customer pain points
1)How to verify the correctness of the application processing of the new core system。
2)How to test the new core system in a full range and variety。
3)How to carry out rapid regression test when interface changes and version upgrades iterate frequently / Solution 1. How to verify the correctness of the application processing of the new core system? 1.Analyze the original message and playback message at the field level, and import the relevant field information into the database. 2. The correctness of the new core application processing is verified by comparing the return code and return field of the old and new core systems. 2. How to conduct a full range and full variety of comprehensive tests on the core system? Adopt the method of small batch iteration.First, select the top 50 transactions with a high proportion of transactions to test, verify their success rate, and gradually increase the number of transactions iteratively under the condition of ensuring the success rate until the whole transaction is covered. 3. How to implement rapid regression testing with frequent interface changes and fast version iteration and update? The method of interface configuration management is adopted.The configuration file of the relevant interface will be created. Here, the method of batch import of the interface is adopted. When the interface is changed, only the configuration file corresponding to the relevant interface needs to be changed to realize the rapid regression test. 4. How to realize asynchronous simulation of core business system test environment?
1.Migrate full desensitization of production AS400 data to AS400 tester (data backup).
2.Migrate all data to x86 simulation distributed database.
3.Conduct full comparison of data.
4.Capture the production environment message, upload it to the desensitization server, and desensitize the message.
5.Transmit the desensitized message to parallel simulator 1.
6.Replay the desensitization message to AS400 tester, capture the packet at the same time, and transmit the message to parallel Simulator 2.
7.Analyze and process the message of parallel Simulator 2 and replay it to the test server.
8.Field level comparison and analysis of messages.
5. How to realize the parallel simulation of core business system production environment?
1.Online migration of production AS400 data to AS400 tester (data backup).
2.Migrate data to x86 simulation distributed database.
3.Deploy A1 in a quasi production drill environment.
4.Capture the production message through the switch image.
5.Process the production environment message in real time.
6.Replay the message to the new core system in real time.
7.Compare and analyze the messages of the old and new core systems at the field level.
/ Program advantages
1)Reduce the technical requirements of operators and personnel investment:
All operations are carried out by the front-end interface, and the regression test can be carried out only by simple configuration,
which does not have high technical requirements for operators.
2)Make automated regression testing easier:
In the case of frequent interface changes and rapid version iteration and upgrade, we adopt interface configuration management.
Each time the interface changes or version upgrades, we only need to simply adapt the configuration file to test again, making regression testing more convenient.
3)High message reusability:
Once the message is captured, the regression test verification of multiple version upgrade iterations can be carried out
4)Save time and cost:
Greatly shorten the test cycle and improve the test efficiency
2)How to test the new core system in a full range and variety。
3)How to carry out rapid regression test when interface changes and version upgrades iterate frequently / Solution 1. How to verify the correctness of the application processing of the new core system? 1.Analyze the original message and playback message at the field level, and import the relevant field information into the database. 2. The correctness of the new core application processing is verified by comparing the return code and return field of the old and new core systems. 2. How to conduct a full range and full variety of comprehensive tests on the core system? Adopt the method of small batch iteration.First, select the top 50 transactions with a high proportion of transactions to test, verify their success rate, and gradually increase the number of transactions iteratively under the condition of ensuring the success rate until the whole transaction is covered. 3. How to implement rapid regression testing with frequent interface changes and fast version iteration and update? The method of interface configuration management is adopted.The configuration file of the relevant interface will be created. Here, the method of batch import of the interface is adopted. When the interface is changed, only the configuration file corresponding to the relevant interface needs to be changed to realize the rapid regression test. 4. How to realize asynchronous simulation of core business system test environment?



Background Interface Display


