I have this issue on AWS on some servers. Whenever I run sudo the terminal is stuck doing seemingly nothing, until it finally spits out this error message. My terminal looks like this:

ubuntu@(none):~$ sudo true
sudo: unable to resolve host (none)

What can I do to solve it?

share|improve this question
2  
Please post the contents of /etc/hostname and /etc/hosts. – arrange Aug 31 '11 at 20:24
39  
I recommend against closing this question as too localized. There are many users who may mistakenly think they've put one name in their hosts file but put in a different name instead, especially since on many networks, computers are similarly named. This question (and answer) would show up when someone searches with that problem, and the answer would prompt them to check for such discrepancies, even though the exact misspelling would be different. – Eliah Kagan Aug 18 '12 at 11:09
6  
make sure your hostname same with hosts. e.g. the hostname is ubuntu-pc and hosts is ubuntu-pc must be same. – Muhammad Sholihin Apr 1 '13 at 8:17
1  
I ran into this today. The problem was that what I had in hostname wasn't in /etc/hosts. To wit: $ hostname => 'mybox' $ grep 'mybox' /etc/hosts => 192.168.1.2 mybox.example.com. I needed to add 'mybox' after my domain name in /etc/hosts => 192.168.1.2 mybox.example.com mybox – William Mar 25 '17 at 21:16
1  
I can't post an answer because this question is protected and I don't have enough reputation here. In my case, I solved the problem by restarting network-manager: sudo /etc/init.d/network-manager restart. However, I'm wondering why in the first place sudo wastes time waiting for network-related stuff. Shouldn't sudo work without problems when network is not available? – bli Jul 31 '17 at 8:45

18 Answers 18

up vote 746 down vote accepted

Two things to check (assuming your machine is called my-machine, you can change this as appropriate):

  1. That the /etc/hostname file contains just the name of the machine.

  2. That /etc/hosts has an entry for localhost. It should have something like:

     127.0.0.1    localhost.localdomain localhost
     127.0.1.1    my-machine
    

If either of these files aren't correct (since you can't sudo), you may have to reboot the machine into recovery mode and make the modifications, then reboot to your usual environment.

share|improve this answer
25  
The hostname will not change until you reboot. If you wish to change it without rebooting the machine then follow the above steps and after that run:- "sudo hostname my-machine" to see if this has worked run "sudo hostname" It will show your machine's host name. This method maybe used as a temporary method to change hostname also. after a restart, the value from the /etc/hostname file is used. – Yashvit May 5 '13 at 15:22
1  
Note: since you can't sudo to begin with, it is difficult to edit those files. My solution was I was somehow able to sudo visudo and change #%admin ALL=(ALL) ALL to %admin ALL=NOPASSWD: ALL , then reboot, and sudo su -, edit those files, set/correct hostname, reboot again, and everything worked. – Ian M Jul 17 '16 at 20:19
4  
I'm using Linux Subsystem in Windows and I faced this problem. After following your answer, It has been resolved. – amarVashishth Dec 28 '16 at 15:43
    
@amarVashishth can you list the steps taken? I am facing the same problem. I tried what the answer tells to do but failed – Meet Taraviya Feb 3 '17 at 12:08
1  
you may also need to add ::1 localhost to /etc/hosts (this is the IPv6 version of 127.0.0.1, aka the loopback address) – Woodrow Barlow Sep 27 '17 at 14:51

Edit /etc/hosts and append your new hostname to the 127.0.0.1 line (or create a new line if you prefer that).

Mine looks like:

127.0.0.1       localhost localhost.localdomain penguin

# The following lines are desirable for IPv6 capable hosts
::1     ip6-localhost ip6-loopback
fe00::0 ip6-localnet
ff00::0 ip6-mcastprefix
ff02::1 ip6-allnodes
ff02::2 ip6-allrouters

Replace penguin in the above example by your new hostname as stated in the /etc/hostname file.

share|improve this answer
2  
How can he edit the /etc/hosts file if he can't sudo? Unless he created a rood account with a password (bad idea) – Dennis Mar 29 '15 at 22:31
7  
@Dennis You can still execute sudo even if that message is displayed. IIRC you still have to enter your password at each invocation though. If this does not work, you can reboot into the recovery console and apply the changes. A root account with password is discouraged. – Lekensteyn Mar 29 '15 at 23:31

Note, this is an answer to this question which has been merged with this one.

Your hostname (dave00-G31M-ES2L) is not represented in /etc/hosts. Add an L to this line:

127.0.1.1   dave00-G31M-ES2

So it becomes:

127.0.1.1   dave00-G31M-ES2L

In order to accomplish this, open a console (press Ctrl+Alt+T) and type:

sudo gedit /etc/hosts

Add the letter L as mentioned, save and exit.

share|improve this answer
9  
Remember! Use sudoedit (or sudo -e). To specify preferred editor, use the EDITOR environment variable (eg. export EDITOR=vim) as it creates an offline copy for editing and then cleanly overwrites after editing. – Jan Sep 26 '14 at 13:54
2  
And there is another one here who suggest sudo when there is no longer sudo. sudo doesn't work, sir. sudo: unable to resolve host ... – Green Jul 12 '16 at 20:18
    
@Green: No sudo? The error message you mention comes from the sudo command. Perhaps you meant something different? – Thor Jul 14 '16 at 13:49
    
@Green sudo works just fine. It just can't store any state (i.e. as Lekensteyn said elsewhere you have to enter your password every time). – Wlerin Jul 28 '16 at 19:16
4  
If you have this issue on W10's Bash and came to this question from google, this is the answer that worked for me. I changed the 127.0.0.1 to look like "127.0.0.1 localhost DESKTOP-SLQK4CV" (by doing "sudo vim /etc/hosts" (quick tip for vim newbies: press i before typing to switch to insert mode, press esc to exit that, write ":wq" to save and exit or ":q!" to exit without saving), in my case sudo worked but just said that it can't connect to DESKTOP-SLQK4CV) and it started worked for me. – Ave Aug 5 '16 at 2:13

Add your hostname to /etc/hosts like so:

echo $(hostname -I | cut -d\  -f1) $(hostname) | sudo tee -a /etc/hosts
share|improve this answer
6  
This was a brilliant answer that didnt get enough credit – Anekdotin Oct 17 '16 at 1:55
    
Nice one! Agreed it doesn't get enough rep love. :-) – Peter K. Oct 13 '17 at 12:39

I had this issue when I was using ubuntu on a VPS. I solved it editing /etc/hosts file.

run this command:

sudo nano /etc/hosts

and then add:

127.0.0.1   localhost.localdomain localhost
127.0.1.1   ubuntu

I hope that will solve your issue :)

PS: Remember to reboot your computer!

share|improve this answer
    
Also, see if your device name (printed on the Terminal title bar after the @ sign) matches the name on the second line of the hosts file ("ubuntu" in Luca's example). The first line may also be just "localhost". – Waldir Leoncio Oct 26 '13 at 14:33
2  
Remember! Use sudoedit (or sudo -e). To specify preferred editor, use the EDITOR environment variable (eg. export EDITOR=vim) as it creates an offline copy for editing and then cleanly overwrites after editing. – Jan Sep 26 '14 at 13:55

I was having the same issue even though the hostname in my /etc/hostname file and /etc/hosts file matched.

My hostname was "staging_1". It turns out that you can't have an underscore in your hostname, which is why I was getting this error. Changing the underscore to a hyphen fixed my problem.

share|improve this answer

In AWS, go to your vpc and turn on "DNS Hostnames".

share|improve this answer
2  
Welcome to askubuntu! Can you expand a bit on this? It's not abundantly clear what you mean (at least to me).. – Elder Geek Jan 15 '15 at 13:56
    
This may not be relevant to the question, but it greatly helped me. Thank you! – clayzermk1 Apr 10 '15 at 17:35
1  
This was the answer that helped me. Amazon AWS changed since the last time I looked at it. VPCs have DNS options, and they need to be turned on before any DNS resolution will work. – Chris Moore Jul 1 '15 at 1:16
    
the Enable DNS Hostnames option can be found (for example) in the right-click menu of the vpc entry – Matteo Scotuzzi Jan 3 '16 at 17:32
    
There is no DNS Hostnames – Green Jul 12 '16 at 20:11

In my case it was the problem, I changed the hostname to man because I wanted to know if there are some parameters you can use on hostname. Instead it changed my hostname to man and I always got the same message like you

sudo: unable to resolve host (none)

after changing the hostname back to `localhost everything worked fine again

hostname localhost
share|improve this answer

I encountered this same error message. I think this discussion thread at AWS Developer Forums is a better solution:

"Go the the VPC management console, select the VPC, click on Actions, select Edit DNS Hostnames and select Yes."

https://forums.aws.amazon.com/thread.jspa?messageID=699718

share|improve this answer

Some terminal emulators will not update prompt with the correct hostname until you close and restart the emulator (lxterminal, I'm talking to you).

I spent 30min fighting with this error after editing my hostname and hosts files and running sudo service hostname restart until I ran sudo hostname and saw that the hostname was the new value, even though the prompt was showning the old value.

share|improve this answer
    
The issue isn't the terminal emulators, it's the shells that have cached the value. – Mark Stosberg Feb 16 '17 at 19:53

Sorry I can't help you much but, since it says "can't resolve host" try running:

hostname

And see if the output is the hostname of the machine. If not, the problem is the host configuration, not sudo.

share|improve this answer
    
I did that got the hostname of my machine. I also have entry in /etc/hosts. I'm still getting the error. – chandresh May 6 '17 at 13:22
    
Did you reboot the machine after changing the hosts file? – darent May 7 '17 at 18:52
    
No. Though my IT team said there was issue with lock file as well as gpg key with puppetlabs. Now, it got resolved without restart.. – chandresh May 8 '17 at 9:28

OP wrote:

It was all in /etc/hostname. On two of our sick servers it looked like this:

ubuntu@(none):~$ cat /etc/hostname
linux-web-n ip-10-128-##-##

While on a server without this issue we had:

ubuntu@ip-10-128-##-###:~$ cat /etc/hostname
ip-10-128-##-###

Removed the linux-web-n portion, rebooted and everything was fine.

share|improve this answer

The symptom given in the question may correlate strongly with this more specific problem:

$ hostname --fqdn
hostname: Temporary failure in name resolution

There are different ways that this could be resolved, one of which is to add your hostname as localhost in /etc/hosts (as shown in several other answers). This may be the right thing to do in general, but it isn't the only possible resolution.

A "fully qualified domain name" may be supplied by an external DNS server or similar (if such is available on your network). In this case, sudo will not complain, despite the missing entry in /etc/hosts.


Note: sudo attempts to dereference the hostname, even though it isn't necessarily required, due to optional capabilities in the sudoers file. See sudo command trying to search for hostname.

As long as the delay isn't too long, this error message is typically harmless.

share|improve this answer

you might be getting an error if your hosts or hostname file contain illegal characters. Only these symbols are permitted: a-z, A-Z, 0-9

share|improve this answer

I had the same problem. I solved it by editing the /etc/hosts and /etc/hostname files... on the /etc/hosts file, just edit the top part as shown below.

#vi /etc/hosts
    127.0.0.1   localhost
    127.0.1.1   localhost  myhostname




#vi /etc/hostname
    myhostname
share|improve this answer
    
127.0.1.1 localhost myhostname or 127.0.1.1 myhostname? – Mostafa Ahangarha Mar 26 '16 at 18:43
1  
How can you edit /etc/hosts without sudo. sudo doesn't work sudo: unable to resolve host ... – Green Jul 12 '16 at 20:02

if you can't sudo you CAN log in as root via su. IE: su root (in an x-term). then give the root password when prompted, then you can edit the files with nano. The root password in 'buntu is the same as the password you would use for sudo.

share|improve this answer
2  
The root password in Ubuntu is not the same password for sudo. root is its own account, which doesn't have a password set by default. – TheWanderer Aug 24 '16 at 19:38

I had this same problem! I changed my VPS's name through the online admin control panel which did not change the machine name in the hosts file All I did was run:

sudo nano /etc/hosts

Then I edited it from this:

127.0.1.1 Megabyte Megabyte
127.0.0.1 localhost

To this:

127.0.1.1 Debian Debian
127.0.0.1 localhost

and that fixed my error! Hope this helped!

share|improve this answer

If you are using Vagrant, then login into the guest and run apt-get --no-install-recommends install virtualbox-guest-utils

share|improve this answer

protected by Community Mar 31 '17 at 3:42

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.