Our latest news
In any company it is essential to have the maximum information from its systems. Unfortunately, most companies are left with the availability of their infrastructure. It is important to know the availability of the infrastructure, but this is not enough to know the exact state of the systems.
When we talk about monitoring, we must first make a distinction between availability and performance. Monitoring availability is by no means sufficient, since it is possible to have all our infrastructure available, but the performance is not adequate, which can cause the total inability to work for our users, a dissatisfaction of them, or even a loss of customers. It is therefore essential to have both availability and performance as objectives.
Since it is necessary to include performance as one of the objectives of any department or group dedicated to monitoring, the following aspects come to mind when monitoring our infrastructure:
- What should we monitor?
- How should we monitor?
There exist 3 monitoring approaches:
- System monitoring, where either physical or virtual systems are monitored, such as the application layers (DB, application servers…).
- Device monitoring, where the communication device is monitored, including the possibility to monitor machines (and their usage) as if they were another device.
- Log-based monitoring (Splunk type), where we try to centralize all the logs in order to dispose of all the information centralized.
Usually, companies have one of the approaches mentioned above, the most common being system monitoring. Here, it is important to repeat the point mentioned above. If our systems monitoring system only shows us availability information, we should think about adding (we do not have to replace it, since monitoring solutions can be easily integrated) a solution that complements us from a performance point of view.
In order to obtain the most complete vision possible, it should be important to complement the three approaches mentioned above, having information from all the layers that make up my infrastructure, which include the systems part (from the end user experience, through APM, heterogeneous databases (SQL and NoSQL), operating systems, virtualization and storage), the communications part (networks, switches, routers, QoS, network usage by applications) and of course, centralization of logs to have the most complete information possible on the state of our infrastructure.