One aspect of running a dedicated server or VPS that many beginners overlook when they get started is monitoring the system for problems. Many users upgrade to their first dedicated server after having their websites hosted in a shared web hosting environment where the management of the server is handled by the hosting company’s sysadmins. For some, the number of elements that they need to monitor and manage can come as quite a surprise. So let’s have a look at some of the common things to watch out for.
Updates To The Max
Firstly we have updates. Operating systems and software packages receive a number of updates in their lifetime, ranging from adding new features to fixing bugs and vulnerabilities. Whilst on a desktop computer updates will usually be automatically performed for you, in server environments the updates normally need to be applied manually. Keeping the software up to date is an important part of protecting it from malicious attackers, so either configuring automatic updates or regularly checking for updates and applying them yourself is crucial.
Memory, Disk Space and CPU Load
Secondly we have the server’s resources. When we talk about resources, we mean the amount of memory in the system, disk space and the CPU load. Running out of memory can cause unpredictable results as different software packages may crash or be killed by the operating system to keep things running, though ultimately the end result of running out of memory is your server going offline. Running out of disk space can cause very similar symptoms, as applications may crash when they are unable to write data to the disk. A high CPU load, on the other hand, is manifested by a server running slowly as applications take a long time to respond to requests, unless the cause of the load is found and stopped then the CPU load will often spiral upwards until the system crashes or locks up. Generally the solution here is to have a monitoring system look at the available system resources and alert you automatically as disk space runs out, RAM fills up or the CPU load increases.
The next thing to monitor is your server’s RAID array, if it has one. RAID is often used in servers in order to provide redundancy in case of a hard disk failure. Two hard disks will be configured into a RAID 1 array, whereby one disk will mirror the data on the other. In the case of one of the two disks failing the system will continue to run on the other disk seamlessly. Should the remaining disk fail though, your system will go offline. So it’s important to have monitoring in place to alert you once a disk in your RAID array has failed so that you can quickly arrange for it to be replaced.
Another important thing to monitor are the system processes that run on your server. For example, if your server is running as a web server you’ll want your monitoring to make sure that the website is accessible. Similarly, if your server is performing duties as an FTP server or email server then you’ll want your monitoring to be checking that those servers are accessible. Pro-active monitoring and alerting here can mean identifying and fixing a problem before your users even notice it was offline.
Are You Online?
The final thing to monitor is that your server is online. Sometimes something can happen that take your server offline or make it appear to be offline. Having a remote computer ping your server and alert you when it doesn’t get a response can help you respond quickly to a potential problem. When you do get an alert it’s always worth testing with a third party ping test such as isup.me or similar before rushing to raise a support ticket.
Pro-active monitoring and alerting can make server management much simpler and make your life administering the server easier, meaning that the time invested in getting it set up will be time well spent.
Never miss another post — Sign up for the 100TB newsletter and follow us on Facebook and Twitter.
Originally published at blog.100tb.com.