Listen to traffic and increase the loyalty of bank customers: myth or reality

Dedicated to all who use banking services.

“Please wait, the information is being loaded,” says the girl at the other end. When you call the bank, at such times, you mentally prepare to expect an unpredictable amount of time on the line. But here passes some miserable 5 minutes and they return to you with an answer. The funny thing is that they then still ask to evaluate the quality of service. And you have the tenth missed call on the second line and a couple of SMS with curses. Well, who does not happen? Suspensions or delays in the operation of banking software were, are and will be. The code is imperfect, and the normal operation of the software cannot be guaranteed by its manufacturer. Of course, each software manufacturer seeks to eliminate bugs as they are discovered, but not everyone succeeds in doing this quickly. Processing user data, increasing the load on the business application, updating the operating system generates unpredictable errors. The article will tell you how we sniffed user traffic and helped a large Russian bank to identify software problems.

Any graduate of the courses of the Codecademy portal developer knows that the real data of real users in a real environment reveal errors in situations that are impossible in the conditions of test benches. The business application support service usually learns about these hard-to-find bugs from the users themselves. Not the most pleasant situation.

Even the presence of monitoring with synthetic transactions, control of code calls within a business application, and even more of infrastructure monitoring, will not always help in understanding the source of the problem. And with the help of these tools, implicit problems cannot be presented to the developer who supports the business application. To solve the problem, we used Micro Focus (formerly HP Enterprise) Real User Monitoring (RUM) user traffic analysis tool. Real case studies and other details below.
')

Why Micro Focus RUM?

Micro Focus RUM is part of the Micro Focus Operations Bridge (OpsBridge) monitoring solution. This integrated solution includes modules for monitoring business processes, synthetic and infrastructure monitoring, which the bank has already used. Micro Focus RUM became an additional building block and helped to look at business applications from the other side. It works on the principle of a sniffer: parses network traffic and reveals real user transactions in it. Important killerficha - the ability to write plugins for listening to traffic that is not supported out of the box.

What are business applications?

We set up Micro Focus RUM to monitor the Internet bank, front-office system, ABS with processing and interaction of the bank with international payment systems. About each application in order.

Internet bank

The bank set us the task of early detection of problems in the work of users in the Internet bank. Prior to the introduction of Micro Focus RUM, the bank’s technical support learned from the untranslatable play on words when the client called the call center, and didn’t recognize any of them at all. Sometimes customers preferred not to report a problem, and silently sent to the bank, discontent rays used alternative methods of making payments (other competing banks) or tried to perform the operation later. Of course, such silence led to the loss of profits of the bank, lack of information about the problems and image losses.

Micro Focus RUM allows you to set predefined client actions in the settings and control the passage according to user scenarios: opening a deposit, processing an application for a loan or credit card, money transfers. When implementing the solution, we set up monitoring of the success of these scenarios, counting the number of users in the scenarios, calculating the cash equivalent, including the size of the deposit, transfer or the requested amount in the loan.

As a result, the bank received a tool for early diagnosis and detection of abnormal operation of the system before the user accesses. The adjusted linking of events between RUM, synthetic and infrastructural monitoring allowed us to see the cause of the problem and reduce the time for its elimination. Profit!

Front Office System

Front-office system is used by bank specialists for customer service: providing information about debts, making transfers, arranging deposits, loans and credit cards. The most common calls to the support service were associated with a long wait for the response of the system interface. Waiting for the interface could go up to 10 minutes. The contractor who serviced the system shrugged, advised to contact the sport lot or provide something more substantial than user complaints. And the bank, in fact, had nothing to show.

After the introduction of Micro Focus RUM, data appeared on the problematic modules of the front-office system, reports on the passage of users on the interface screens. A total of 12 such transaction algorithms were set up to place orders for bank products: a cash loan, a mortgage, a credit card, a deposit, payment transactions, and other significant transactions. We adjusted accounting of transactions by macroregions and cities of presence of offices, the bank was able to view user transactions in real time and receive regular reports. The most important thing is that the contractor now receives complete data for making improvements.

ABS

The bank uses an ABS system that works with a closed database without the ability to make requests from outside. Such, it turns out, still occur. Processing interacts with it using a fully proprietary protocol, which has resulted in an obstacle for setting up monitoring. The captain of evidence suggests that without performing direct queries to the database, we will not be able to receive complete information about the state of its components.

The possibility of using Micro Focus RUM in this situation is not obvious. We scratched our turnips and found a way out - listening to the traffic between the application and the database. Next, after analyzing the traffic, we figured out the types of database queries and developed a plugin for parsing it.

After setting up monitoring with Micro Focus RUM tools, the bank now has the opportunity to follow the work of the ABS and make suggestions for improvements to the contractor.

Monitoring of interaction with payment systems

The bank interacts with international payment systems using a standard protocol (according to RFC 4949). Each card transaction is a payment transaction with a call to an external service (Multicard).

In this case, Micro Focus RUM came to the rescue with control of data exchange between payment systems and ATMs with payment terminals. As with working with ABS, we analyzed and developed mechanisms for parsing this type of traffic. The result was the ability to track the status of transactions in real time and receive regular reports.

As a result, the bank received detailed data on the functioning of applications, monitoring the impact of changes in them, and diagnostic tools for transferring problems to the contractor. Now we are planning to expand to other applications.

Author of the article: Anton Kasimov , architect of monitoring systems, Technoserv company.