How to monitor the work of the business process

There can be a hundred or so monitoring systems in the world. It could be and cloud systems, on-premise, and on payment basis, free, for the network, infrastructure and so on. Among them there are those that support the creation of service-resource models. These are tree-type catalog, to which the elements of the business system are tied to the nodes: web servers, databases, application servers, switches and many other terrible words. Each child element affects the parent. For example, if a server threshold is exceeded on a certain server, an event is created (let’s say critical criticality - red), which affects the objects higher in the structure of the service.

Many organizations link the availability of the business system to the availability of the business process. If I had considered SLA, it would have turned out that the availability of the IT system had dropped to zero (at the time of exceeding the memory threshold on some server there), and the business process had stopped. But this is not so? For communication between business and IT, you need to somehow evaluate the availability of services and business processes. There is a very simple way: there was an incident in the company - start counting out the inaccessibility. Finished work - stop the counter.

The real picture of the business process can be provided only by metrics that characterize the success and availability of business process steps. As a result, two isolated systems are obtained with their own events and accessibility, but in a single interface. If we see that one of the steps in the business process does not work that way, it's an excuse to look at related IT systems. We consider unreliable the impact of the system on the process, but for the shift attendant or the owner of the process / system left the opportunity to view this connection for diagnostic purposes.

What is needed in general to implement such an algorithm:

- determine the composition of the company's processes (what exactly we want to control);

- determine the impact of key IT metrics on these processes (for example, the availability of any channel, without which 50% of business does not work and the retail director calls);

- decompose the impact on individual systems, and them on infrastructure and so on;

- implement the specified two-contour model - business process monitoring and key transaction events plus diagnostic information from IT systems and infrastructure.