What are the main differences between remote monitoring Linux and Windows servers?
Q: If I need to monitor a Linux server, should I choose entirely different software for that?
A: The primary difference between Linux and Windows servers is in the choice of software and/or protocols used to retrieve required data.
Whereas Windows offers WMI as universal scripting tool allowing to get and set a variety of system parameters (such as file attributes, general system parameters etc), Linux can use an SNMP agent as a provider of system data and uses a number of tools providing the same functionality (vmstat, stat, iotop, data in /proc filesystem). Thus, in case of Windows you could use WMI, whereas you should use SNMP Custom monitor or Script and/or SSH monitor to get data from a Linux server. There are 4 predefined hardware performance monitors that collect their data via SSH protocol, so they are especially easy to use if the monitored host is Linux. You can monitor several CPU load metrics, various memory usage parameters, free or used disk space, or varied process characteristics of any SSH-enabled host.
All the software pieces useful in case of Windows have the Linux alternatives and vice versa. For example, you could use WMI and programs like Speed Fan to watch and control CPU, motherboard and hard disk drives temperature. In case of Linux, use smartctl and lm-sensors for that purpose, both provide their data via SNMP protocol. WMI can be used to access bandwidth data of network interfaces in Windows; SNMP can be used for exactly the same in Linux. That goes for all the system health and performance tuning parameters such as CPU load, memory available, hard disk space, programs/services running state, uptime and so on.
If the same protocol is used, monitor parameters choice is in most cases the same, regardless of the operating system type and version. Parameters can differ if the actual server hardware set and/or location differs (i.e. other ping response time etc. should be used).
