What is Swap Space in Linux Systems?

Imagine you are running a Linux workstation. You’ve got a few browser tabs open, a virtual machine running in the background, a large dataset loading into a Python script, and you’re writing documentation notes in a markdown editor, all at once.

Suddenly, your system becomes sluggish. Your applications take longer to respond, and your terminal is lagging,  due to the lack of available memory.

This is not a hypothetical situation. In fact, this can happen to anyone at any time with little warning. 

The question is, how do you deal with this?

Swap space, a dedicated area on your disk that acts as virtual memory when your system’s physical RAM is full, is the solution to sluggish and slow-responding systems. 

In this article, I will discuss what swap space is, its types, and how it stacks against comparable solutions. Let’s start with a quick overview of swap space.

What is Swap Space in an Operating System?

Swap space is a designated area on your system’s storage (either a hard drive or SSD) that the operating system uses as an extension of physical memory (RAM). 

When your system’s RAM is full, the OS moves inactive or less-used data from RAM into swap space, freeing up RAM for active processes. This process is called swapping or paging.

In my years of experience, I have lost the count of times when swap space became the savior of system performance. If you are monitoring a production server and see memory usage climbing toward 100 percent, then you should know that the kernel is preparing to start moving less frequently accessed pages from RAM to swap space.

In my experience, the most common misconception I have seen is that swap space is some kind of RAM replacement or RAM accelerator, which, by now, you understood is absolutely not meant to act as the main memory.

Note that swap is much slower than RAM, so if your system is constantly using swap, you’ll notice a big performance hit. This is called thrashing.

A few other misconceptions about swap space that I have come across are:

• Disabling swap improves performance: Without swap, your system may crash or kill processes when RAM runs out
• Swap should always be double your RAM: That’s outdated advice. The right size depends on your workload and system needs.
• Swap is useless if you have lots of RAM: Even with plenty of RAM, swap helps optimize memory usage and can prevent out-of-memory errors

Types of Swap Space in Linux

In Linux, you can implement swap in two main ways: swap partitions and swap files. 

Let us discuss each of these:

Swap Partitions: The Traditional Approach

A swap partition is a dedicated section on the hard drive or SSD that is formatted specifically for use as swap space. It functions as an extension of the system’s physical memory (RAM) to provide additional virtual memory when the RAM becomes full. 

The swap partition is isolated from other data, ensuring efficient and reliable management of virtual memory. Its primary role is to temporarily hold data when the RAM usage is 100 percent. 

Advantages of Swap Partitions

• Raw performance: Swap partitions have slightly better I/O performance since there’s no filesystem overhead
• Kernel efficiency: The kernel can directly manage the partition without going through filesystem layers
• Reliability: Swap partitions are less prone to corruption issues due to its separate, dedicated position on the physical media
• Simplicity: Once created, swap partitions require minimal maintenance
• Boot-time availability: Typically, swap partitions are always accessible during system startup. This means the system can always have support in terms of RAM usage

Disadvantages of Swap Partitions

• Fixed size: You can’t easily resize without repartitioning, which requires downtime or rebooting the system
• Disk space commitment: The space on the disk is permanently allocated, even when not needed
• Complex setup: Requires partition table modifications that I prefer to avoid on production systems
• Inflexibility: Moving to a different disk requires significant reconfiguration of the target disk’s filesystem

Swap Files: The Flexible Alternative

A swap file is simply a regular file created within the existing filesystem and configured as swap space. I’ve seen system administrators increasingly adopt this approach, especially in virtualized environments and cloud deployments where access to physical disk is not always available.

Advantages of Swap Files

Let us see some of the advantages of swap files.

• Dynamic sizing: You can easily resize swap by recreating the file without touching the partition tables
• No repartitioning: Swap files can be added to any existing filesystem with enough available space
• Multiple swap areas: You can create several swap files across different filesystems for load distribution
• Easy management: Standard file operations, backup policies, and migration processes apply on swap files
• Quick deployment: Perfect for emergency situations when swap space is required immediately

Disadvantages of Swap Files

• Slight performance overhead: The filesystem layer adds minimal but measurable latency
• Fragmentation concerns: On heavily used filesystems, the swap file might become fragmented
• Filesystem dependency: If the underlying filesystem has issues, swap can become unavailable
• Security considerations: Swap files are regular files that can be accidentally accessed if permissions are misconfigured, causing security issues.

The following table summarizes the difference between a swap partition and swap files.

Now on to the real question: when to choose one over the other. 

I always prefer swap partitions when:

• Setting up dedicated servers where performance is critical
• Working with systems that have predictable, stable memory requirements
• Managing bare-metal installations where I need to control the entire disk layout
• Dealing with high-I/O workloads where every microsecond matters

Alternatively, I choose swap files when:

• Working in cloud environments where flexibility is paramount
• Managing virtual machines that might need memory adjustments
• Dealing with systems where I can’t easily repartition disks
• Setting up development or testing environments that require frequent changes

Swap Space vs. Virtual Memory: What’s the Difference?

One of the most frequent questions I noticed system administrators get from junior administrators is about the relationship between virtual memory and swap space. 

I believe that understanding this hierarchy is crucial for effective system management.

What is Virtual Memory?

Virtual memory is a system that allows your computer to use more memory than is physically available by combining RAM and swap space. In other words, virtual memory allows the operating system to use disk storage as an extension of RAM, enabling memory-intensive applications to run even when physical memory is limited.

How Does Virtual Memory Use Swap Space?

The OS divides the memory into fixed-size blocks called pages. In addition, the OS keeps track of which pages are in RAM and which are in swap. 

When your system runs out of RAM, the OS moves less-used data (pages) from RAM to swap space. This process is called paging. 

If a program needs data that’s been swapped out, the OS brings it back into RAM, possibly swapping something else out to make room. This is called a page fault.

Now, if the system is constantly swapping pages in and out, performance drops sharply. This is called memory overflow or thrashing. If the system needs to swap in and out pages constantly, it is a sign that you need more RAM or optimize your workload in-memory data distribution.

How Swap Space Works

As someone who has closely worked with sysadmins, I’ve seen how swap space acts as a safety net for Linux systems when memory usage starts going into red. 

The kernel, the core part of the operating system, is responsible for managing memory through swapping or paging. When the system detects that memory resources are running low, it initiates a series of steps to free up RAM:

• Monitoring Memory Usage: The kernel continuously tracks the amount of free RAM and other memory states.
• Identifying Pages to Move: It identifies less-active or unused memory pages that can be safely moved to swap space without affecting ongoing processes.
• Swapping Out Pages: These pages are temporarily moved/written to the swap space on disk, freeing up RAM for more recent data storage requirements.
• Swapping In Pages: When a process requires data stored in swap that is no longer in RAM, the kernel reads these pages back into RAM.

However, you need to keep in mind that when many pages are swapped frequently, you will notice the following effects:

• Performance Degradation: When the system is swapping heavily, applications slow down dramatically 
• Disk I/O Bottlenecks: Heavy swap usage increases disk activity, which can bottleneck the system, especially on slower disks or in cloud environments with I/O limits 
• System Responsiveness: The system may become sluggish or unresponsive as it spends more time swapping than running applications 
• Thrashing: If both RAM and swap are full, the system can enter a state called thrashing, where it spends most of its time swapping pages in and out, rather than doing useful work.

How Much Swap Space Do You Need?

In my interactions with senior system administrators, this is the most frequently asked system management question I get. The interesting thing here is that the answer has evolved significantly over my career. 

During the late 1990s, RAM was expensive and limited. A typical server had 128MB-512MB of RAM, and therefore, the old rule of swap was to have at least 2x disk space of the RAM. 

However, in the modern systems with 32GB, 64GB, or 1TB of RAM, the 2x rule is overkill and can even be counterproductive because of the following reasons:

• RAM is abundant: Modern systems have 100x more RAM than systems from 20 years ago
• Workloads have changed: Applications are designed for high-memory environments with appropriate memory management components
• Storage implications: 128GB of swap space on a 64GB RAM system is rarely justified
• Performance expectations: Users expect systems to perform well, not just survive

Therefore, consider the following factors to calculate how much swap space you need:

• Workload type: Memory-intensive applications (databases, scientific computing) may need more swap as a buffer.
• Available RAM: The more RAM you have, the less swap you typically need.
• Hibernation needs: If you use hibernation, you need at least as much swap as your RAM to store the system state

The following table illustrates how much swap you need.

How to Check Swap Space in Linux

I always recommend regularly checking your swap usage, especially if you suspect performance issues.

The following are some of the common commands you can use to check swap space.

Use swapon –show

If you want to display detailed information about active swap areas, including name, type, size, and used space, execute the following command:

# swapon –show

Check /proc/swaps

If you want to know the details about all swap areas currently configured, check the /proc/swaps file. Use the cat command to view the file and see which swap areas are in use on your system:

# cat /proc/swaps

Use free -h for a Complete Memory Snapshot

If you want to display the total, used, free, shared, buffer/cache, and swap memory in a human-readable format, run the following command:

# free -h

Monitor Swap Usage with top or htop

If you want to display real-time system info, including swap usage, towards the top of the output, run the top command:

# top

Now, if you want an enhanced version of top with a more user-friendly interface, use the htop command:

# htop

The following table summarizes these common swap commands:

How to Add or Remove Swap Space

Managing swap space in Linux involves adding new swap areas, resizing existing ones, or disabling swap when necessary.

You can either make temporary changes (swap space for the current session) or permanent changes (same after reboots).

Follow the steps below to add swap space.

Step #1: Check Existing Swap Space

Start by checking the current swap space and identifying the swap partition. 

# swapon –show

Step #2: Turn Off the Swap

Before resizing, I recommend turning off the swap space.

# sudo swapoff -v /[swap_path]

Replace [swap_path] with the partition path.

Step #3: Resize the Partition

Once you have disabled the swap space, use a partitioning utility such as fdisk or parted to decrease or increase the swap partition.

# sudo fdisk /dev/sdX

Important: Resizing partitions with fdisk can cause data loss. Back up data first and verify disk paths with lsblk.

Step #4: Update the Partition Table

Once the swap partition is resized, update the partition table with the following command:

# sudo mkswap /[swap_path]

Step #5: Enable the Swap Space

Now, enable the resized swap space.

# sudo swapon /[swap_path]

Replace [swap_path] with the appropriate partition path.

Now, you need to remember that increasing the size of an existing swap partition is not possible if the maximum storage space is already occupied. In such cases, you need to create a new swap file.

Advantages of Using Swap Space

The following are some of the advantages of swap space that I have noticed:

Provides an Overflow Buffer

When RAM is full, swap acts as an overflow, allowing the system to keep running by moving inactive data to disk. This prevents out-of-memory errors and keeps critical applications running.  

Enables Hibernation

Hibernation requires saving the entire contents of RAM to disk. Swap space is used to store this data, making it possible to power down and resume exactly where you left off. 

Prevents System Crashes

Without swap, running out of RAM can cause the kernel to kill processes or the system may crash. Swap provides a buffer that helps prevent these situations, especially under unexpected memory requirement spikes.

Improves Multitasking

By freeing up RAM for active processes, swap allows more applications to run concurrently, improving multitasking and overall system stability

Extends Usable RAM for Resource-Limited Devices

For older hardware or devices with limited RAM, swap provides an affordable way to extend usable memory capacity, improving usability and application compatibility.

Helps in System Maintenance and Troubleshooting

Swap can be useful during system upgrades or troubleshooting, providing extra space for processes that need temporary buffers or disk-based memory for diagnostics.

Disadvantages of Swap Space

While swap is essential, it’s not without drawbacks. These are some of the drawbacks that I have noticed:

Slower Than RAM

Accessing swap is much slower than RAM. If your system is using swap heavily, you’ll notice sluggish performance, increased latency, and longer application response times. 

SSD Wear

On SSDs, frequent swap usage can contribute to wear, as SSDs have a limited number of write cycles. Modern SSDs have good wear leveling, but it’s still best to minimize unnecessary swap writes.

Poor Swap Management Can Lead to Sluggish Systems

If swap is overused or not sized/tuned properly, the system can become unresponsive, especially if it enters a state called thrashing, where it spends more time swapping than doing useful work 

Disk I/O Bottlenecks

Heavy swap usage increases disk I/O, which can slow down other processes and degrade overall system performance 

Possible Misconception of Adequate Memory

Heavy swap usage might give a false impression that the system has enough RAM because the system appears functional, but in reality, performance degrades because of the slow disk-based memory.

Conclusion

Swap space is a vital part of Linux memory management. It acts as a safety net when RAM is exhausted, supports hibernation, and helps prevent crashes. However, because swap is much slower than RAM and can impact SSD lifespan, it’s important to size and tune swap appropriately for your workload.

Always remember that swap is not a RAM replacement. 

FAQs

How does the file system impact swap space performance?

The file system type affects how efficiently swap files are accessed. While swap partitions bypass the file system for faster performance, swap files depend on the underlying file system, which can introduce slight overhead. 

Can swap space wear out my hard disk over time?

Yes, excessive swap usage on traditional hard disks (HDDs) can lead to faster wear, especially during sustained memory pressure. On SSDs, although performance is better, frequent writes to swap can reduce the drive’s lifespan. It’s important to allocate enough RAM and optimize applications to minimize reliance on swap.

Does the amount of RAM affect whether I need swap space?

Absolutely. The more RAM your system has, the less likely it is to use swap space under normal conditions. However, swap is still recommended even on high-RAM systems as a safety net for rare memory spikes or when running memory-intensive applications like virtual machines or databases.

Is swap space used to store data permanently?

No, swap space is not meant for permanent data storage. It temporarily holds inactive memory pages from RAM to free up space for active processes. Data in swap is volatile and cleared on reboot; it doesn’t replace or supplement your hard disk’s long-term storage function.