In this article, we will tell you how to monitor CPU temperature in remote servers and computers that run Windows.
It makes sense to closely watch your servers’ and workstations’ critical health parameters (for example, the temperatures of components that may overheat, especially if it’s a hot summer). A failure of a CPU, video adapter, or hard drive will bring the entire server or workstation to its knees and may even result in a data loss. The best way to prevent a situation like that is to eliminate the conditions that can cause it. Make sure that the hardware is duly maintained and any failing cooling systems are repaired or replaced in good time, provide adequate air conditioning, and so on.
In most cases, computer components do not overheat out of the blue. If you keep tabs on the situation, you can take the necessary measures before anything bad happens. So make sure that you monitor the temperatures and fan speeds in computers. Thankfully, nowadays most motherboards have sensors for that purpose. The BIOS built-in hardware monitoring features are often useless, but you can easily find many good third-party applications to monitor CPU temperatures and fan speeds. The most popular of them are probably SpeedFan, CoreTemp, Open Hardware Monitor, Hmonitor, and Real Temp.
The problem is, all of these apps are intended for the local computer, so you will have to instruct multiple users to check temperatures and fan speeds on a regular basis. In case of a server, you are out of luck, because you need to be logged in to use an app like that.
But what if you need to implement centralized monitoring of all computers in your network, including automatic notifications in a critical situation? Actually, you can build an enterprise monitoring system based on the open-source software Zabbix and write scripts to parse Open Hardware Monitor’s console output. This approach can be useful if your company or organization already uses Zabbix and has a specialist who can configure it. But there is a much easier approach that allows you to save time and effort, and also doesn’t require any special skills.
10-Strike Network Monitor Pro can monitor the temperatures of CPUs, hard drives, and video adapters, as well as fan speeds and other parameters supported by the motherboard sensors. All you need to do is install it on one of your servers, and also install Network Monitor’s agents on the computers that you want to monitor. After that, Network Monitor will scan the network, find any available hosts, and add them to the monitoring list. You need to create the check “Temperature and other sensors” for the computers on which Network Monitor’s agent is installed, and it will run at regular intervals. The monitoring core will analyze the values received from each agent and compare them with the threshold values. If the threshold is exceeded, Network Monitor can notify you about that by sending an SMS text message, email message, etc.
But it is not enough to be informed about something bad after the fact. It is much better to foresee a critical situation by analyzing accumulated data. 10-Strike Network Monitor Pro stores each received value in its database. It can plot graphs and generate reports based on collected data. Say, you can take a look at a fan speed graph, see that the RPM has been decreasing for the last few weeks (maybe due to bearing degradation), and replace the fan in good time. If the CPU fan speed is normal, but the CPU temperature has been slowly increasing for the last few weeks, it’s probably time to replace the CPU thermal grease.
Monitoring Setup
It is easy to create a CPU temperature check in the network monitoring system. You need to do the following:
1) Install "10-Strike Network Monitor Pro"
2) Install Network Monitor’s agent on each host you want to monitor. The agent is a Windows service that will read the current temperature value and send it to the monitoring core.
3) Launch Network Monitor and run a network scan, or manually add the hosts.
4) Select the host in the tree (at the left), and then select “Add check” in its context menu.
5) In the “Check parameters” dialog box, select the check type: “Temperature and other sensors.” After that, click the “Get” button (at the right, below the “Motherboard sensors” pane).
6) Find the hardware name in the sensor list and expand its subtree. Select the parameter you want to monitor and set the alarm triggering conditions (for example, “The check is successful if the sensor value is less than 50”).
7) Set the check start parameters or keep the default ones. Set the notification options, and save the changes.
As soon as you add a new check, it will start collecting data. You can watch the data collection process by switching to the “Monitored parameter” tab (click the tab at the bottom of the window). Network Monitor will display a temperature graph.
Conclusion
As you can see, it takes only 7 easy steps to implement a 24/7 monitoring of your servers and workstations’ critical health parameters, and it doesn’t involve any highly technical, hard-to-understand methods.
10-Strike Network Monitor Pro allows you to build a distributed monitoring system. It means you can use it in a large company or organization that has multiple remote offices, each one with its own local network. The advantage of having a distributed system is that all the monitoring servers in remote networks will send the data they’ve collected to the centralized database, and all the settings and check results will be stored in one place. Network administrators will probably love this solution, as they won’t have to set up port tunneling on a remote network’s router each time they need to configure monitoring for that network.