The performance index is also a key factor with the increase of heating enterprises' scale and years of operation. In particular, the seasonal characteristics of the heating industry are obvious. Many heating enterprises have a very short charging period. In many cities, heating enterprises need to handle thousands or even tens of thousands of charges a day in peak hours, plus integrated services such as network access, shutdown reporting, reheating, switch on and off, and query, The business system is heavily loaded. Many charging business halls have long queues in the busy season. In this case, if the software performance and stability have problems, it is easy to have small group events. Therefore, the charging software has also made some definitions on the performance indicators of the Heating Charging Management System (200000 households, system response requirements under the condition of ordinary desktop solid state disk): normal page query operation: 0.2 seconds. Charging business operation: 2 seconds (query/charging/refund/shopping cart/stop reporting/reduction/invoicing, etc.). General report query: 6 seconds. Complex report query: 15 seconds. Special report and batch operation: set up requirements according to the actual situation. The function and performance of the product quality Heating Charge Management System is visible.

This is also an important step. As we continue to deepen our data analysis, we find that the data migration work of Datong Thermal is much more complex than we thought. This complexity is not the huge data magnitude mentioned above, but the non-standard data. We learned that some functions of the original charging system in the application process can not meet the business needs, so people who operate the software often use an unconventional operation to achieve a certain business, such as directly modifying the heating fee receivable if the recovery cannot be achieved after the suspension is reported, directly modifying the standing book if the area settlement is not correct, and Director Zhang also said.

Compare and assess. Management target: The toll management system of the non local group version is designed and developed for the cross regional group heating enterprises, and builds a comprehensive information platform around customer, area and cost management. It supports different management modes of multiple locations and multiple subsidiaries to achieve strict separation of data between subsidiaries, while the group headquarters can supervise and review subsidiaries, and achieve the integration of all business data, solving the problems of repeated investment and data sharing of group companies. According to the needs of local policies, the subsidiaries in each region independently improve the customer's heat use files, clarify the heating area and status, open up multiple service channels such as WeChat, Alipay, banks, counters, and so on, and internally link the finance and offline mobile terminal switch, check, and clear up debts, so as to achieve the goal of each department.

Hainan Easy to charge Toll software The premise is that we are worried about slow acceptance, so we should set aside more time for simulation exercises. Since leaders attach importance to training, we will certainly give full cooperation. I said that training can be carried out at any time. On August 5, according to the requirements of Director Xu, I drew up a training schedule: the training of the charging system was divided into two batches, one for the business personnel of the customer service center and the business hall, and the other for the branch companies. One batch of training is divided into 8 sessions, one session every half day, with 105 participants, mainly explaining 9 functions and 59 functional modules of the system; There are four training sessions in the second batch, which are also held every half day. The total number of participants is 56, mainly explaining 32 functional modules of the nine major functions of the system. The training of the customer service system is also the same. In general, the training intensity is quite high, especially for our engineers. On August 10, the training was in full swing. Our chief engineer had to take a break because of his sore throat. It turned out that not all challenges were so easy to succeed. After all, we were human beings. Without any loudspeakers on the scene, could we speak to so many people every day without interruption without pain? But the training couldn't stop, so I took the place of two trainings.

In order to seek market survival opportunities. In fact, your concentration and professional ability determine whether you can survive this winter. Secondly, let's take a look at the development of the heating industry. With the promotion of the national coal to gas land conversion, the energy cost of many heating enterprises has risen sharply. In addition, the real estate has gradually cooled down, and the income of supporting fees has also become less and less. In addition, the local government finance is not loose, and financial subsidies are not available. This situation has led to many heating enterprises operating difficulties, It will also lead to difficulty in collecting payment from IT vendors. The changes in the external environment have also prompted heating enterprises to turn to refined operation, focusing on energy conservation and consumption reduction and improving management efficiency. On the other hand, many private capital have transformation needs, and they are also consciously acquiring some heating enterprises.

There has never been such a big fight on other projects, because the magnitude of data cannot be reached. This time, we have organized a special team to be responsible for the performance test and optimization of Datong Thermal. The test is divided into performance troubleshooting stage and stress test stage. Performance troubleshooting stage: test all the functions of the system according to the test cases, find out the functions that lead to the increase of system resources and slow response, record them, and optimize the performance problems exposed in this stage. Stress test phase: perform concurrent stress test on common functions. In this phase, we focus on several main performance indicators: response time, throughput, and system resource consumption. First, the common functions that need stress testing are determined and the test scripts are sorted out. Determine the test concurrency of 5,10,15,30,50 according to the actual use of the customer, and simulate 5010015030500 people using the system at the same time. Lock the functions with performance problems according to the concurrency test results, and determine how to allocate the resources of each virtual machine and how many Tomcat clusters to create. On site training At the beginning of August, Director Xu called me to the office to discuss the system training.