How to improve supervision and performance of business IT systems

Companies often encounter problems with the performance of their IT systems, which affect the smooth running of their business activities. The reason? The monitoring and reporting tools only partially process the information arising from monitoring the network and applications' performance.

Corporate information systems are increasingly reliant on hybrid models that mix traditional infrastructure with cloud-based solutions. Big data and AIOps make it possible to have a supervision model that effectively monitors performance: global visibility, predictability and adaptability to the business.

Global visibility across all of your IT performance

Why? For a comprehensive analysis of the origins of incidents in order to identify the causes and take corrective actions more quickly. To do so, all the layers that make up the information system need to be analyzed: infrastructure, network, applications and user experience.

How? By aggregating and correlating all the information provided by the monitoring tools in real time. This makes it possible to centralize all the events related to the information system and their impact on the company's business.

What is the outcome for the support service? Increased productivity in operational management, meaning:

  • Fewer incident tickets: for the same problem, the company avoids multiple tickets being created. Currently, 20 to 30% of the incidents created automatically based on monitoring information are redundant
  • Time saved on resolving incidents: due to the rapid identification of the root causes of incidents
80%
of incidents that cause problems with performance are not effectively corrected due to a lack of information

 

Cross-functional visibility thus helps to remove the disparities that exist between all of the monitoring tools, which are generally "mono-service." Currently, it is difficult to identify the root cause of breakdowns or loss of performance. The IT ecosystem consists of different interconnected and interdependent services, relying on proliferation of monitoring tools and dashboards with fragmented visibility. This situation does not help to detect and correct incidents, especially in the shadowy areas not covered by the tools.

Predictability of problems

Why? To take a proactive approach to incident management and protect against problems by deploying preventive actions.

How? By adopting an Artificial Intelligence Operations systems model, or AIOps. This makes it possible to store, combine and analyze data collected in real time by the monitoring tools deployed by the company and the correlation rules (such as Machine Learning), which generate “smart alerts.”

What is the outcome for the support service? Anticipation of technical issues, recurrent or not, which could harm the company's business:

  • Fewer incidents. The company avoids any potential deterioration or even saturation
  • Identification of any loss of performance, access to applications, the cloud, voice, etc. by exploiting and correlating the available information about how they are functioning. The company avoids any deterioration even before users become aware of it
  • Identification of areas for improvement. Using history and evolution of the data, mature trends are created, making it possible to determine how the IT needs to be adapted
70%
of IT organizations learn about performance problems from end users
Source: Gartner

 

By making problems predictable, the company avoids business-critical situations. The history of events allows it to analyze the origins of problems and prevent them from recurring. Currently, 70% of IT organizations learn about performance problems from end users after they appear, and by the time the cause has been found, the situation has very often returned to normal, making problems difficult to manage.

Adapted supervision model

Why? To take into consideration the company's business activities when monitoring different sites, services and business applications…and best meet the requirements of the business. For example, office applications are more critical for administrative and commercial workstations than in a store.

How? By adapting the AIOps model by considering the different levels of criticality of the monitored sites and services. According to the impact of incidents on the company's business, there would be a more relevant classification and prioritization of alerts.

What is the outcome for the support service? Improved decision making for the operational management of the IT system:

  • An information system at the service of the business: monitoring the performance of the network and ensuring that business-critical applications are working properly, especially if the business is seasonal
  • An information system that can adapt to needs: responding, for example, to an increase in an activity

By integrating its business needs, the company can customize its supervision or monitoring mode and optimize the management of its critical IT services:

  • An e-commerce site for a company that sells online
  • A stock management application for a retailer
  • A journey planning application for technicians in the field, etc.

Conclusion

By adopting these three principles (global visibility, predictability of problems and adapted supervision model), it should be easier to monitor the entire IT performance, thus ensuring business continuity for the company.

Mathieu Chapuis
Mathieu Chapuis

As a Business Developer in the Key Accounts Department for monitoring and AIOps services, I support our account teams and customers with their innovation and deployment projects for these high added-value services.