One of the easiest way of increasing the responsiveness of your server and guarding against out-of-memory errors in applications is to add some swap space. In this guide, we will cover how to add a swap file to an Ubuntu 16.04 server.
Swap is an area on a hard drive that has been designated as a place where the operating system can temporarily store data that it can no longer hold in RAM. Basically, this gives you the ability to increase the amount of information that your server can keep in its working “memory”, with some caveats. The swap space on the hard drive will be used mainly when there is no longer sufficient space in RAM to hold in-use application data.
The information written to disk will be significantly slower than information kept in RAM, but the operating system will prefer to keep running application data in memory and use swap for the older data. Overall, having swap space as a fall back for when your system’s RAM is depleted can be a good safety net against out-of-memory exceptions on systems with non-SSD storage available.
Before we begin, we can check if the system already has some swap space available. It is possible to have multiple swap files or swap partitions, but generally one should be enough.
We can see if the system has any configured swap by typing:
If you don’t get back any output, this means your system does not have swap space available currently.
You can verify that there is no active swap using the free
utility:
Output total used free shared buff/cache available
Mem: 488M 36M 104M 652K 348M 426M
Swap: 0B 0B 0B
As you can see in the “Swap” row of the output, no swap is active on the system.
The most common way of allocating space for swap is to use a separate partition devoted to the task. However, altering the partitioning scheme is not always possible. We can just as easily create a swap file that resides on an existing partition.
Before we do this, we should check the current disk usage by typing:
OutputFilesystem Size Used Avail Use% Mounted on
udev 238M 0 238M 0% /dev
tmpfs 49M 624K 49M 2% /run
/dev/vda1 20G 1.1G 18G 6% /
tmpfs 245M 0 245M 0% /dev/shm
tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs 245M 0 245M 0% /sys/fs/cgroup
tmpfs 49M 0 49M 0% /run/user/1001
The device under /dev
is our disk in this case. We have plenty of space available in this example (only 1.1G used). Your usage will probably be different.
Although there are many opinions about the appropriate size of a swap space, it really depends on your personal preferences and your application requirements. Generally, an amount equal to or double the amount of RAM on your system is a good starting point. Another good rule of thumb is that anything over 4G of swap is probably unnecessary if you are just using it as a RAM fallback.
Now that we know our available hard drive space, we can go about creating a swap file within our filesystem. We will create a file of the swap size that we want called swapfile
in our root (/) directory.
The best way of creating a swap file is with the fallocate
program. This command creates a file of a preallocated size instantly.
Since the server in our example has 512MB of RAM, we will create a 1 Gigabyte file in this guide. Adjust this to meet the needs of your own server:
We can verify that the correct amount of space was reserved by typing:
Our file has been created with the correct amount of space set aside.
Now that we have a file of the correct size available, we need to actually turn this into swap space.
First, we need to lock down the permissions of the file so that only the users with root
privileges can read the contents. This prevents normal users from being able to access the file, which would have significant security implications.
Make the file only accessible to root
by typing:
Verify the permissions change by typing:
Output-rw------- 1 root root 1.0G Apr 25 11:14 /swapfile
As you can see, only the root user has the read and write flags enabled.
We can now mark the file as swap space by typing:
OutputSetting up swapspace version 1, size = 1024 MiB (1073737728 bytes)
no label, UUID=6e965805-2ab9-450f-aed6-577e74089dbf
After marking the file, we can enable the swap file, allowing our system to start utilizing it:
We can verify that the swap is available by typing:
OutputNAME TYPE SIZE USED PRIO
/swapfile file 1024M 0B -1
We can check the output of the free
utility again to corroborate our findings:
Output total used free shared buff/cache available
Mem: 488M 37M 96M 652K 354M 425M
Swap: 1.0G 0B 1.0G
Our swap has been set up successfully and our operating system will begin to use it as necessary.
Our recent changes have enabled the swap file for the current session. However, if we reboot, the server will not retain the swap settings automatically. We can change this by adding the swap file to our /etc/fstab
file.
Back up the /etc/fstab
file in case anything goes wrong:
You can add the swap file information to the end of your /etc/fstab
file by typing:
There are a few options that you can configure that will have an impact on your system’s performance when dealing with swap.
The swappiness
parameter configures how often your system swaps data out of RAM to the swap space. This is a value between 0 and 100 that represents a percentage.
With values close to zero, the kernel will not swap data to the disk unless absolutely necessary. Remember, interactions with the swap file are “expensive” in that they take a lot longer than interactions with RAM and they can cause a significant reduction in performance. Telling the system not to rely on the swap much will generally make your system faster.
Values that are closer to 100 will try to put more data into swap in an effort to keep more RAM space free. Depending on your applications’ memory profile or what you are using your server for, this might be better in some cases.
We can see the current swappiness value by typing:
Output60
For a Desktop, a swappiness setting of 60 is not a bad value. For a server, you might want to move it closer to 0.
We can set the swappiness to a different value by using the sysctl
command.
For instance, to set the swappiness to 10, we could type:
Outputvm.swappiness = 10
This setting will persist until the next reboot. We can set this value automatically at restart by adding the line to our /etc/sysctl.conf
file:
At the bottom, you can add:
vm.swappiness=10
Save and close the file when you are finished.
Another related value that you might want to modify is the vfs_cache_pressure
. This setting configures how much the system will choose to cache inode and dentry information over other data.
Basically, this is access data about the filesystem. This is generally very costly to look up and very frequently requested, so it’s an excellent thing for your system to cache. You can see the current value by querying the proc
filesystem again:
Output100
As it is currently configured, our system removes inode information from the cache too quickly. We can set this to a more conservative setting like 50 by typing:
Outputvm.vfs_cache_pressure = 50
Again, this is only valid for our current session. We can change that by adding it to our configuration file like we did with our swappiness setting:
At the bottom, add the line that specifies your new value:
vm.vfs_cache_pressure=50
Save and close the file when you are finished.
Following the steps in this guide will give you some breathing room in cases that would otherwise lead to out-of-memory exceptions. Swap space can be incredibly useful in avoiding some of these common problems.
If you are running into OOM (out of memory) errors, or if you find that your system is unable to use the applications you need, the best solution is to optimize your application configurations or upgrade your server.
Thanks for learning with the DigitalOcean Community. Check out our offerings for compute, storage, networking, and managed databases.
This textbox defaults to using Markdown to format your answer.
You can type !ref in this text area to quickly search our full set of tutorials, documentation & marketplace offerings and insert the link!
I’m not too sure I understand the warning about having a swap file being “dangerous” for your hardware. If that is true, wouldn’t it be a problem if my application writes several MB of logs every day, enlarge it’s database by many MB every day… etc.
Is there is a difference between a swap file or other files that make them so dangerous?
If swap is heavily used on a daily basis it experiences way more writes per day than any combination of your db a app writes, so SSDs degrade faster. Basically, swap is an emergency tool in your belt. If your server starts to use it - get ready to refactor your app or to upgrade your droplet soon. If swap is used all the time, 20% or more - you should upgrade your droplet immediately. Used swap means you’re running out of memory.
Is there an easy way to know how active swapping is over time? I always get a little bit of swap used over time, but that generally does not tell me how much is used in 1h, 1 day, 1 month…
Google sysstat and sar. This tool might help.
vmstat can also tell you if memory is moving in and out of swap
Hi - thanks for a great article. I have configured swap on ubuntu 16.04 with an Nginx web server, Wordpress underneath, it has plenty of spare memory on a $20 server but it seems to be all eaten up with 1.2GB of buff/cache, and is entering swap even with swappiness set to 10. Of course I would prefer swap space is only used in an emergency. Any thoughts - is this Ubuntu or Nginx eating it up? Here is the extract from Top.
… and from “free”
I’m pretty sure it’s MySQL. Incorrect db buffers/cache settings can really eat a lot of memory. Use mysqltuner to check it.
Yeah … did that at the beginning, the suggested mysql settings made no difference to this.
As a side note, certain buffers may be saved to swap and never accessed again in which case it stays in the swap memory and does not do anything one way or the other until the applications “using” (i.e. “not using”) that buffer quits. Thus, it is normal that you would eventually see large buffer/cache memory available and swap being used.
Thanks. Reply to your WARNING - generally, swap on SSD might be ‘evil’ but it’s necessary to run some apps or to be able to upgrade some apps on basic 512 MB RAM droplet (in my case, I’ve tried to upgrade MariaDB; without swap it just failed and I was unable to run it again ever, with swap it upgraded like charm and runs smoothly).
I’ll leave this here as it can help someone in the future: Heads up on “Make the Swap File Permanent” step.
I had to destoy one droplet because it was entering on “linux emergency mode” after finishing up this tutorial and I couldn’t find the reason. As I created a brand new droplet and finished this tutorial again, it was entering on emergency mode just like the last one.
After some minutes thinking what would be the problem, I realized that I got wrong the step “You can add the swap file information to the end of your /etc/fstab file by typing:”
echo '/swapfile none swap sw 0 0' | sudo tee -a /etc/fstab
The command above should be inserted and commited on the bash line command, and I have inserted it as is on the last line of the file
/etc/fstab
.A very simple mistake, but I’m a begginer here and I got used to
sudo nano
files and insert the codes by hand on the other 10 previous tutorials from DO that I’ve consumed. I hope this save someone droplet one day…On the permanent part; you could also add the swapon and sysctl commands to your .bashrc.
I also found a solution to seeing ‘Device or resource busy’ everytime:
The article is spot on, and works like charm. Thanks. :-)
I think to say SSDs will degrade (fail) quicker was true with earlier SSDs. Newer ones don’t seem to have this issue, and by the time it eventually failed, the SSD would be outdated and need replacing anyhow.
This isn’t to say you should use swap on a regular basis to replace physical memory.
Using Ubuntu 16.04.1 x64 w/ 512MB RAM … without swap there was not enough Mem to run apt-get. Nice straight forward tutorial. Enjoyed learning about swappiness and cache pressure :)
Can you add all links for each version of Ubuntu for swap? https://www.digitalocean.com/community/tutorials/how-to-add-swap-on-ubuntu-12-04 https://www.digitalocean.com/community/tutorials/how-to-add-swap-on-ubuntu-14-04 https://www.digitalocean.com/community/tutorials/how-to-add-swap-space-on-ubuntu-16-04
“We can verify that the swap is available by typing:” “sudo swapon --show”
When I do that to verify, it gives no output, and the output of free -h is still the same as before:
This is on Ubuntu 16.04.
@b9ac4679230fd17: Looks like you missed the first
swapon
command that actually activates the new area as swap.After the
mkswap
command, useswapon
without the--show
flag to activate the file you created:Afterwards, you should be able to see the new swap space.
Thank you for the effort you put into this. So well written, and helped me learn more about the topic!
Last time I read one of these guides, there was no warning about swap. I’ve not noticed any ill effects from using swap other than that my stuff no longer crashes, so is it worth doubling or quadrupling how much I pay…?
I have just set up a fresh new Droplet with Ubuntu 16.04 and when I run
I get an error:
Any idea what’s wrong here? I’ve successfully used this tutorial for other droplets in the past. Now it seems not to work any longer and I have absolutely no idea why?!
@manuelbieh Hey there. I’m only able to reproduce the problem you are describing when I copy and paste your example. If I type it in directly, it works fine.
It appears that the character you’re using for the dash in
-l
is not the same character that it’s expecting. There are a number of different characters that look the same. It looks like the character it is expecting is the hyphen-minus sign, while your keyboard is supplying a regular hyphen. You can view the slight difference here. If you use the hyphen-minus symbol within the command instead of the hyphen, it should work correctly.On another note, we don’t recommend setting up swap on DigitalOcean Droplets because it can increase the likelihood of hardware failures for both you and anyone else on the same physical server. If at all possible, it is always a better idea to increase the amount of RAM instead.
Hope that helps.
Indeed! After typing it (instead of copy+paste it) it worked. I saved this article as PDF and copy+pasted it from there. Seems like ‘Save as PDF’ smuggled some invisible characters into it. Thanks!
I’ve made this into a quicky .sh script, with a one line copy and paste into the terminal to download it, run it, and then cleanup. Figured I’d share it with you guys since this is my 7th time doing this.
Kinda seems silly that this isn’t done when a droplet is created.
https://github.com/spikeon/digitalocean_swapfile
i used your script and this created 2gb swap space in just a sec. Does this make swap permanent a described in tutorial. Thanks a lot. :)
After creating swap how to cleaup this from server. I have used sftp but it said permission denied to delete swapfile in root.
Just found the swapfile in root i was deleteing is the actual swap created not .sh script. Sorry. Does your script automatically clean itself
The script should clean itself. it would be located in the home folder at .swap (~/.swap) if it failed to cleanup after itself.
I should’ve mentioned that I changed it to 2gb, my bad.
I wrote it to be run as root or sudo, probably forgot to mention that as well.
It does do the extra steps to make the swapfile permanent, don’t want it going away after a system reboot.
If you need to delete the swapfile you’ll need to run
swapoff
first to ensure it’s not being used, and thensudo rm
, however the swapfile in root is there for a reason, if you want the swapfile you probably don’t want to delete it.This comment has been deleted
If u are newbie and bought 512mb set swapfile 4GB
Here is a simple one-liner to replace all that stuff:
Why bother with all of that plumbing - if there is already a module for that?)
Great description! I followed it in detail but the boot time in Ubuntu 16.04.3 increased from 35s to 380s. Is there a way to reduce the boot time? Thanks. Styl.
@styl It sounds like you might have a timeout occurring. You may need to double check your entries to make sure they’re correct. Viewing the boot logs will probably help you find the culprit too.
Another option is to do swaping with a zram device. It is basically a compressed ramdisk for swapping.
See https://www.kernel.org/doc/Documentation/blockdev/zram.txt
It can help a lot if the swapped out data is good compressible.
I followed this guide to add swap space in my machine. Everything goes fine till I reach this step:
Here, it throws this error:
swapon: /swapfile: swapon failed: Input/output error
Has anyone faced this error? How to deal with it?
The error message
swapon: /swapfile: swapon failed: Input/output error
when trying to enable the swap file can be caused by several issues. Let’s troubleshoot this step by step:Correct Swap File Creation: Firstly, ensure that the swap file was created correctly. The typical steps are:
sudo fallocate -l size /swapfile
, wheresize
is the desired size of your swap file (e.g.,1G
for 1 GB).sudo chmod 600 /swapfile
to make it readable and writable only by the root user.sudo mkswap /swapfile
.If any of these steps were missed or executed incorrectly, it could result in the error you’re seeing.
Filesystem Support: Make sure your filesystem supports the swap file. Some filesystems, particularly network file systems or some special filesystem types, might not support swap files properly.
Use of
fallocate
vsdd
: If you usedfallocate
to create your swap file, some filesystems might not correctly allocate the space. Althoughfallocate
is faster, it might not work on all filesystems. In such cases, you can usedd
to create a swap file:bashCopy code
sudo dd if=/dev/zero of=/swapfile bs=1M count=1024 sudo chmod 600 /swapfile sudo mkswap /swapfile
This example creates a 1 GB swap file (
count=1024
with a block size of 1M). Adjust thecount
value for a different size.Filesystem Errors: Check for filesystem errors. Sometimes, input/output errors can be a result of filesystem corruption or issues. You can check and repair your filesystem with tools like
fsck
, but be cautious as runningfsck
on a mounted filesystem can cause data loss. It’s safer to run it from a live CD/USB or in single-user mode.Kernel Messages: Look at the kernel log messages for any clues. Use
dmesg | tail
immediately after you get the error. This might provide more information on what is going wrong.Swap Support in Kernel: Ensure that your kernel supports swap files. This is usually not an issue with standard distributions and kernels, but it’s something to consider if you’re using a custom kernel.
Swap File Integrity: Check the integrity of the swap file. An interrupted or incomplete
fallocate
ordd
process could result in a corrupted swap file.Reboot the System: In some rare cases, a system reboot can resolve temporary issues that might be causing the swap file to be unreadable.
Alternative Swap Method: If nothing works, consider creating a swap partition instead of a swap file, especially if you have unallocated space on your disk.
If you continue to face issues, it might be helpful to provide more context or error logs for a more specific diagnosis.
Top notch. Well ordered and easy to follow. Good explanations.
Hi, I’m new here. I use ubuntu 16.04 I made a swap space, looks everything ok to me. But when I open Gparted, I only see /dev/sda1. From used space I understand, that I created a swap space, but I don’t see seperate line with swap? Thax in advance.
When you create a swap space on a Linux system like Ubuntu, there are two main ways to do it: using a swap partition or a swap file. The method you choose can affect how the swap space is displayed in tools like GParted.
Swap Partition vs. Swap File:
Checking Your Swap Setup:
swapon --show
orcat /proc/swaps
in the terminal. These commands will display your current swap configurations. If you see a file path (like/swapfile
), it indicates a swap file. If you see a device path (like/dev/sdaX
), it’s a swap partition.Understanding GParted Display:
Benefits of Swap File:
Performance Considerations:
In summary, if you don’t see a separate swap partition in GParted, it’s likely you’ve created a swap file, which is a common and efficient way to add swap space on modern Linux systems.
Very detailed step by step tutorial… perfect!!
Very clear and useful tutorial. My mongorestore command was crashing on my staging server when trying to set it up with production data. After setting up a swap of 2GB (droplet has only 1GB or ram) the import went well. It’s a one time process, with no time pressure, so upgrading the ram just for the import felt like overkill.
I am not able to create the swap file even though I have enough space in the root directory. This is what my
df -h
returns:Can anyone please help? My MySQL db isn’t starting and giving an error of low memory so I wanted to add swap space until I figure out what’s eating up my RAM.
From the
df -h
output you’ve shared, it appears you have sufficient space on your/dev/vda1
partition (8.4 GB available) to create a swap file. If you’re encountering difficulties in creating a swap file on your Ubuntu 16.04 system, there could be several reasons. Let’s troubleshoot this step by step:Check Permissions: Ensure you have the necessary permissions to create and modify files in the root directory. You should be using
sudo
to execute the commands, especially if you’re not logged in as the root user.Correct Swap File Creation Command: The typical command to create a swap file is
sudo fallocate -l size /swapfile
, wheresize
is the size of the swap file you want to create. For example,sudo fallocate -l 1G /swapfile
will create a 1GB swap file. Make sure you’re using the correct command and size.Filesystem Quotas: If you’re using filesystem quotas on your server, they might be preventing you from creating large files even if there appears to be available space. You can check this with
quota -s
.Existing Swap Space: Check if there’s already existing swap space configured using
swapon --show
. If there’s existing swap space, consider whether it’s sufficient or if you still need more.Filesystem Integrity: Run a filesystem check to ensure there’s no corruption or issues. You can use
fsck
, but be cautious with this tool, especially if you’re checking your root filesystem.Creating Swap File with dd: If
fallocate
isn’t working for some reason, you can try usingdd
to create a swap file. For example:sudo dd if=/dev/zero of=/swapfile bs=1G count=1
for a 1GB swap file. However, be very careful with thedd
command as it can cause data loss if used improperly.thx very clear and useful and this solve my problem restoring huge mysql backup file with hundreds of thousand rows (none of online suggestions to increase max_allowed_packets and other time setting works)