Author: Konstantin Boyandin, System Administrator
I have several sites, just like a great number of people has nowadays. As time advances, more and more network services become involved, running a site is not running just a single Web server, there are many network services closely intertwined – email, Web, directory services, database services and so forth.
From this viewpoint, intranets and Web sites tend to have more and more common traits: there are many service depending on one another. Even in a small intranet the number of such services can grow quickly. Smooth operation of every single component of a network facility becomes a must. However, as the number of services increases, the task to control the status of those becomes a hard one.
Time is most precious resource, and the owners of network presence sites, ranging from a personal Web site to large intranets can’t afford manual monitoring of the services. System administrators can’t possible be on watch 24 hours a day, and need for means to monitor the status of all the network services becomes crucial.
Malfunction of even one of services can disrupt the whole operation of network systems. Early warning and the chance to be informed of any unusual state, be it high server load, service suddenly stopped, disk space depleted below acceptable level etc. – this is what is required to have such situation handled quickly.
Human interference is not required in any possible case. In many a case automated means can investigate the resources state and take first action, before human operators can attend to the problem. An example of such a situation could be a denial of service attack (DoS), the modern plague of many small and large service providers. If unusual load on network resources is detected as early as possible, software part can take preliminary measures to reduce possible damage to resources.
Nowadays E-commerce business services monitoring relying on network services can’t underestimate early warning and network monitoring tools. Time and other resources spent to return the failed services back to operation can cost much more than a network monitoring facility, able to watch many types of network activity and inform the personnel by a number of means in case an alert is issued. This is knowledge often learned through an unpleasant experience of having a massive failure of a crucial network service that was run unwatched for a significant amount of time.
To prevent such problems costs much less than to handle them. It never hurts to have good network monitoring tools able to inform you by all means available.