// LEARN LINUX TV • ARTICLE

Using Bad RAM is Risky - Here’s How To Test Yours

ADVERTISEMENT

Bad RAM can cause random crashes, corrupted files, failed builds, and problems that look like software bugs but aren’t. In this video, I explain why you should test your memory, whether you’re building a new PC, setting up a homelab server, or troubleshooting a system that’s acting strangely. Toward the end, I show you how to use MemTest86 to check your RAM for errors before they cause bigger problems.

YouTube player

Join my first-ever live training, where you can learn Nextcloud directly from me. There’s only 30 slots open, so reserve your seat while you can!

What does defective RAM look like?

To get started, let’s talk a bit more about bad memory and the kinds of problems it can cause.

Normally, this is where I’d tell you exactly which error messages to watch out for. The problem is that defective RAM rarely presents itself the same way twice. In fact, over the years I’ve seen memory failures produce all kinds of different symptoms, and no two systems behaved exactly alike. That makes bad RAM surprisingly difficult to recognize.

The best way I can describe defective memory is to call it an “X-File.” It’s a complete mystery until you discover the real cause. You might encounter corrupted files, failed checksum or integrity checks, random crashes, graphical glitches, poor performance, or completely unexplained behavior. One of the most common scenarios I see is a Linux installation failing for no obvious reason. Sometimes the installer reports an error that doesn’t make sense, and after testing the hardware it turns out the real culprit was defective RAM.

The reason the symptoms vary so much is that they depend on when software accesses a defective memory location. If an application never touches the bad area, you may never notice a problem. But the moment it does, the results can be unpredictable. Different operating systems, applications, and workloads exercise memory differently, which is why switching from Windows to Linux—or even installing a different Linux distribution—can sometimes expose a hardware problem that had been hidden for years.

Installing an operating system is actually a great example. Before installation files are written to disk, they’re processed in system memory. If those data structures become corrupted while they’re in RAM, the installer may fail, report checksum errors, or behave in unexpected ways. That’s one reason many people discover defective memory for the first time while installing Linux.

Since I can’t give you a definitive list of symptoms, let me instead offer a few practical guidelines.

First, test the memory in every new or used computer you acquire—especially used systems. Even brand-new hardware can have manufacturing defects, and older systems may have developed memory faults over time. Before you trust a machine with your data, spend the time to verify that its memory is healthy.

Second, make RAM testing part of your troubleshooting routine whenever something doesn’t make sense. If a Linux installer fails without a clear explanation, test the memory. If your system starts crashing randomly, test the memory. If applications are behaving inconsistently, or you’re seeing unexplained checksum or integrity errors, test the memory. It only takes one bad memory cell to send you down a troubleshooting rabbit hole.

Thankfully, testing your RAM is straightforward and doesn’t require much effort. We’ll go through the process in just a moment, but first I want to explain why it’s so important not to ignore defective memory.

The biggest issue is that you can’t fully trust a system with bad RAM. If data becomes corrupted while it’s in memory and is then written to disk, that corruption becomes permanent. A document might become damaged. A database could contain incorrect data. An operating system file could be written incorrectly.

You may also experience kernel panics, application crashes, or seemingly random freezes that make your system unreliable.

Even routine tasks like installing software or applying system updates can become risky. Packages are downloaded correctly, but before they’re installed they’re processed in memory. If defective RAM corrupts that data before it’s written to disk, you could end up with damaged system files and an unstable operating system. At that point, replacing the faulty memory is only part of the solution—you may also need to reinstall or repair the operating system.

So to summarize, defective RAM can prevent Linux from installing successfully, cause mysterious and inconsistent behavior, corrupt your files, and undermine the reliability of your entire system.

Before we move on, I also wanted to mention my brand new Linux Troubleshooting Strategies ebook. Testing your memory is one of the troubleshooting techniques I cover, but it’s just one piece of a much larger toolkit. The ebook walks through proven strategies for diagnosing and solving many of the most common Linux problems, and purchasing a copy is also a great way to support the channel. I really appreciate it.

So now that we’ve covered why memory testing is so important, let’s take a look at how easy it is to do.

Testing your Memory

So how do you actually test your memory?

Fortunately, it’s very simple. All you need is a bootable memory testing utility.

Some Linux distributions actually include one right in the installer menu. Before you start the installation, take a look at the boot menu and see if there’s an option such as “Test Memory” or “Memtest.” If your distribution includes it, that’s the easiest option because you won’t need to download anything else.

Not every distribution offers that feature, though. If yours doesn’t, all you have to do is download a standalone memory testing utility and boot from it.

One option is Rescuezilla, which is primarily a system backup and recovery environment. It also includes a memory testing utility, making it a handy tool to keep around on a USB flash drive for a variety of maintenance and recovery tasks.

Another popular choice is MemTest86, which is dedicated specifically to testing system memory.

Using it couldn’t be much easier. Simply download the ISO image, write it to a USB flash drive just as you would any other bootable image, and boot your computer from it. Once it starts, the memory test begins automatically. From there, it’s mostly a matter of waiting to see whether any errors are detected.

You may notice there are several versions available, including the free and Pro editions of MemTest86, as well as Memtest86+. So what’s the difference?

Originally, Memtest86 was developed by Chris Brady. When development slowed, the project was forked in 2004 to create Memtest86+, which remains an open-source project. Years later, in 2013, PassMark acquired the original MemTest86 and resumed its development. That’s why there are now multiple versions available.

For most people, I recommend the free edition of MemTest86. The Pro version adds features that are useful in enterprise environments, such as more advanced reporting, improved support for identifying which DIMM has failed on supported systems, and additional automation capabilities. It can also reduce testing time on some systems by making better use of available CPU resources. But for the vast majority of home users, the free edition is more than capable of detecting faulty memory.

Memtest86+ is also an excellent choice, especially if you prefer open-source software. Recent versions have added UEFI and Secure Boot support, so they’re much more capable than they were in the past. However, MemTest86 generally supports a wider range of modern memory technologies and hardware features, including more comprehensive ECC memory testing.

My recommendation is to use the free edition of MemTest86. It supports modern hardware extremely well, it’s easy to use, and for most people it provides everything needed to confidently determine whether their system’s memory is healthy.

Diagnosing Memory Issues

Next, let me give you a few additional tips for testing your memory with MemTest86.

The first question most people have is how long they should let the test run.

By default, MemTest86 performs four passes. If all four complete without any errors, there’s a very good chance your memory is healthy. At any point, you can press Esc to stop the test and reboot your system.

In many cases, defective RAM reveals itself fairly quickly—sometimes within the first 10 or 15 minutes. But that’s not always true. Some memory faults are intermittent and only appear under certain conditions. If you’re troubleshooting a system that only crashes occasionally or exhibits seemingly random behavior, it’s worth letting the test run overnight. If you don’t stop it, MemTest86 will simply continue running additional passes.

If the test does detect a problem, it’ll be very obvious. Any memory errors are highlighted in red, so there’s little chance you’ll miss them. If you do see errors, the next step is figuring out which memory module is responsible.

The ideal solution is, of course, to replace the defective RAM—but first you need to identify the bad module. If your system has multiple memory sticks installed, shut it down, remove one module, and run the test again. Repeat the process until you’ve isolated the faulty stick. With a few rounds of testing, you can usually determine exactly which module needs to be replaced.

The Pro edition of MemTest86 can identify the failing DIMM automatically on supported systems, but for most home users, manually testing one module at a time works perfectly well and doesn’t cost anything.

Finally, don’t think of memory testing as something you only do after a problem appears. Any time you purchase a new or used computer, install new RAM, change memory settings in the BIOS, or begin troubleshooting unexplained system instability, running a memory test should be one of your first steps.

If you’re repurposing a used computer for Linux, I especially recommend testing the memory before you spend time configuring everything. Imagine getting your system set up exactly the way you want it, only to discover later that defective RAM corrupted part of the operating system or some of your personal files. At that point, replacing the RAM is only part of the job—you may also need to reinstall the operating system and restore data from backups.

Spending a few hours testing your memory up front can save you many hours of troubleshooting later, and it gives you confidence that the system you’re building on has a solid foundation.

ADVERTISEMENT