Article

Two Types of Host Monitoring on the Network

The server and client-server methods

The host monitoring process is to control the amount of devices connected to the local network, services' efficiency, and to gather the information on average and maximal loading level. This information gives additional opportunities to a system administrator: to react on occurred failures in time, to optimize the hardware's and network device's operation more effectively, and to plan the network upgrades as well.

There are two types of the host monitoring process: the server and client-server methods. The first one is the server-based periodical scanning of the whole range of IP addresses on the network, which allows detecting all the devices connected to it. This kind of scanning usually repeats every 5-10 minutes, which helps to gather the current and up-to-the-minute information. This method does not load the network with a lot of insignificant requests. The second type of the host monitoring process is reading the ARP table cache from network switches. This method is much faster than the previous one and does not require the great amount of requests. But its main drawback is that it cannot help to detect connected hosts, if they do not show any activity.

In comparison with the server host monitoring methods, which gather rather limited amount of data, the client-server methods could constantly read a lot of different computer's parameters, and send them to the server for the further analysis. The main advantage of this kind of host monitoring is that the system administrator is able to build a fail-safe system. If there is no response from the client during the particular period of time, it is polled again. In addition, this method allows generating alerts automatically. In addition to the data on the device availability on the network, the client-server host monitoring method controls all the services that are launched on the remote PC, checks the proper protocols operation, and monitors opened ports. But at the same time, switches and printers are polled by sending ARP requests or just pinging. This method's disadvantage is additional load on the monitoring PC. The RAM load can be rather high, which could be noticeable on old computers especially when other monitoring services are launched at the same time.

To use the obtained information, the mechanisms of visualization and reports' generation are required regardless the chosen host monitoring method. Charts, diagrams, network graphs, and other drawings can be used as the data visualization methods. The diagrams on the amount of working hosts are especially convenient and useful. Besides, usually, the most of host monitoring programs allow the system administrator to view the information in tables, without the graphical representation. Reports can be generated on demand or automatically for a particular period of time.

By the reference to what has been said, it is necessary to keep in mind a few important factors while choosing a good host monitoring program:

  1. An opportunity to organize a fail-safe system.
  2. The graphical interface and tools for generating reports must be available.
  3. The system's operation principle.
  4. Functionality.

Only with the host operation analysis, it is possible to create the effective host monitoring system with minimal load of client PCs and the network itself.

More Stories By Dmitriy Stepanov

Dmitriy Stepanov is a CEO at 10-Strike Software, Network Inventory, Network Monitoring, Bandwidth Monitoring Software developer. 10-Strike Software is a software developing company. They offer their high quality networking products since 1999. 10-Strike Software specializes in producing Windows network software for corporate users.