One of the major credit institutions in Russia experienced problems with its internet banking system, with end users often unable to access the service. The company's IT department could not find reasons for the failure using infrastructure monitoring and was forced to restart the servers to get the system working properly again. They could not find the cause of the "inoperativeness" of the service for a long time.
To find the root of the problem, the bank needed an Application Performance Monitoring (APM) solution, or an APM tool.
This type of tool allows for the collection of infrastructure metrics (CPU, memory, free space, network interface load), as well as monitoring the status of various components of the Service itself (software that enables the final product to work: web servers, content servers, databases, etc.).
Key Challenges
- To solve problems with its internet banking system
- To implement an Application Performance Monitoring (APM) solution
Implemented our solution
ICL Services specialists suggested implementing a Dynatrace solution in the bank, a company that has been at the top of the Gartner Magic Quadrant for eight years in a row.
The customer agreed to use the Dynatrace product in pilot mode. The implementation took place in August 2018. At the planning stage, the ICL Services specialists requested all the necessary resources and indicated all the technical conditions for the work: the list of servers that provide service operation and which will be included in monitoring; ports that must be open for Dynatrace to function; resources required for the Dynatrace product itself.
Since the customer had strict requirements regarding cybersecurity, it was impossible to remotely implement Dynatrace, so it was decided to carry out the implementation on the territory of the Bank. The bank's security service requested a complete description of the work schedule in advance: on which servers the agents will be located, on which ports the communication between agents and the Dynatrace server will work, which accounts and which access rights are necessary for the APM solution to work.
After all the necessary approvals were received and requested technical conditions implemented, the ICL Services specialist was scheduled to leave for the bank's office. The installation and initial configuration of Dynatrace took three days. Additionally, on the second day of work, agents were installed on test servers, since the customer wanted to make sure that Dynatrace does not take up a lot of server resources and does not affect the operation of the Internet Banking service. The next day, when the bank was convinced that Dynatrace had no effect on system performance, the agency's solution was deployed on a "production" circuit - 20 servers that operate the Internet Banking service.
Results
- Thanks to weekly reports on the service as a whole, the business immediately saw how the unavailability of its services affects end users
- For the developers, non-optimised database queries and problem areas in the application code were highlighted
- For the IT department, infrastructure problems + network load and communication between all components of the infrastructure were shown