Mount CIFS Host is down

Question

I have an issue with a mount point that was previously configured. It shows the folder, but the mount is missing and holds "?" values for size, permissions, etc.

So I tried to remount using cifs and the same command from before:

mount -t cifs //nas.domain.local/share /mnt/archive

But I get the error:

Host is down.

If I ping the domain or IP I get a proper resolution and I also connected using smbclient without issue

 ping nas.domain.local
 ping ip
 smbclient //nas.domain.local/share

I looked around, but cant find a solid answer. Any thoughts?

Answer 1

This could also be because of a protocol mismatch. In 2017 Microsoft patched Windows Servers and advised to disable the SMB1 protocol.

From now on, mount.cifs might have problems with the protocol negotiation.

The error displayed is "Host is down.", but when you do debug with:

smbclient -L <server_ip> -U <username> -d 256

you will get the error:

protocol negotiation failed: NT_STATUS_CONNECTION_RESET

To overcome this use mount or smbclient with a protocol specified.

for smbclient: add -m SMB2 (or SMB3 for the newer version of the protocol)

smbclient -L <server_ip> -U <username> -m SMB2

or for mount: add vers=2.0 (or vers=3.0 if you want to use version 3 of the protocol)

mount -t cifs //<server_ip>/<share> /mnt/<mountpoint> -o vers=2.0

Answer 2

On archlinux after a recent package update, I had to add vers=1.0 to my mount options. I'm connecting to an old centos 5 box and up until yesterday I could connect without explicitly stating a version number.

CIFS in linux kernel 4.13 now defaults to SMB 3.0 and in kernel 4.14 it tries 2.1 and higher. See this change log.

Answer 3

USB-stick at Fritz NAS showed "Host Down" for Ubuntu 17.10:

Defining the version (vers=1.0) worked - here's the full string:

sudo mount -t cifs -o vers=1.0,_netdev,username=<user>,password=<pwd>,uid=1000,gid=1000  //192.168.178.1/fritz.nas <local mountpoint>

Answer 4

Similar problem after upgrade to ubuntu 17.10, with an old Buffalo Diskstation. Solved by adding in /etc/fstab the "vers=1.0" option:

//myWDhostname/partage /media/Partage cifs guest,vers=1.0 0 0

Answer 5

Sorry if this is a late response (I realise it's an old thread), however I have just discovered there is another possible reason why mount.cifs would say the host is down.

I have an antivirus with a firewall and even though I set it explicitly to allow "windows file and print sharing" -- a predefined rule, it was still blocking connections. I had that proven by disabling the firewall temporarily. Hope this helps someone, host is down might not mean it's not responding to pings, but could mean it's not responding to authentication attempts.

Answer 6

I received the same error without further ado from a new Samba client, when trying to mount a CIFS SMB network share:

mount error(112): Host is down

Eventually, it turned out I had previously restricted SMB server access to only a limited number of IP addresses by configuring /etc/samba/smb.conf:

# Allow these IP Addresses to connect: 
hosts allow = 127.0.0.1 127.0.1.13 127.0.1.63

# Anything else not allowed is, by default, rejected
hosts deny = ALL

Adding the fixed IP address of the new SMB client solved the issue in this specific case.

Of course, there is a myriad of other reasons why one may receive above-mentioned error.

Answer 7

Same trouble connecting to Synology DiskStation (DSM 4.3).

Using vers=1.0 in the mount options works fine.

Additionally I had to use the option "noperm" because all files wrongly showed as not readable and writable by the owner.

Answer 8

The SMB1 version of the protocol has been deprecated, however this is the default version used in older versions of mount.cifs, e.g. I have this problem with version 6.2.

You can check with: sudo mount.cifs --version

If you try to connect to an SMB3 server using SMB1 protocol, you get the Host is down error.

The workaround, as described by many other answers here, is to specify a different version of the protocol. The following command works for me: sudo mount -t cifs //server.name.or.ip/shares/Public /target/directory -o username=someuser,domain=somedomain,vers=3.0

However, if the server that you are connecting to uses DFS, then you will get the following error instead: mount error(38): Function not implemented. This is because DFS support on SMB3 was only added to the kernel in version 4.11.

You can check your kernel version with uname -a. In my case, it was 3.10 on CentOS7. I followed these instructions to upgrade and now it works.

Answer 9

Same trouble with Fritzbox 7490: mount error(112): Host is down

I didn't used -o vers=XX. As fast as a shark i am, i first tried -o vers=2.0 and failed.
As soon as i used the option -o vers=1.0, everything works fine !

This works for me..

 sudo mount -t cifs -o rw,username=myname_on_the_box,pass\word=mypasswd_on_the_box,vers=1.0 //192.168.1.1/Fritz-nas /media/something/something    

My env:
Client: Ubuntu 17.10 Linux 4.13.0-17-generic #20-Ubuntu SMP x86_64 GNU/Linux
Server: Fritzbox 7490 firmware 6.83.

Answer 10

For me, the mounted cifs share was on a Windows server whose IP address had changed recently, so I could ping the server and resolve its new address, but the mount had not updated itself. By running a lazy unmount and then re-mounting my issue was solved:

umount -l /mnt/share
mount -a

Answer 11

If you're having this problem with a Synology NAS, then check that the vers= option specified to mount and the min/max SMB versions on the NAS are compatible.

Specifically, I'm using vers=2.0, but my Synology Diskstation was triggering the Host is down error. I found a page, Windows 10 access to NAS share. SMB 1.0 and 3.0, on the Synology website that explained how to set the Diskstation to allow SMB v2.0 or newer...

On Synology NAS

• Go to Control Panel-->File Services
• On the SMB/AFP/NFS tab, select Advanced Settings
• Change Maximum SMB protocol to SMB3
• Change Minumum SMB protocol to SMB2 (the page says to use SMB2 with large MTU, but that didn't work for me)

Answer 12

I typically use this type of command to mount a cifs/smb share.

mount -t cifs -o rw,netbiosname=nasserver1,credentials=/etc/user_credentials.txt //192.168.1.11/someshare /mnt

the credentials file looks like so:

username=mydomain\user1
password=somepass

This can also be adapted to an automount setup so the mounting/unmounting can be handled by the system automatically via autofs.

Answer 13

In our case I checked the users login name (of user2) in the AD. There I noticed that the name was starting with an upper case letter and changed it to lower case as it is written in the mount script. Even if we did not touch neither user2 nor the mount script before, suddenly the mount command was successful.

mount --verbose -t cifs //pc/share /my-share -no user=user1,password=pw1 -o uid=user2,gid=group1,dir_mode=0775,file_mode=0664

Answer 14

I also just ran into the problem mentioned after an upgrad to Xubuntu 17.10. I use a Synology DiskStation. What I saw there: In the DiskStation, you can choose which protocols to support. By adding he relevant protocols (up to SBM3) in the advanced options for file services in control panel, you can also solve the problem.

Answer 15

Had a similar problem. The solution for me was on the Windows share server side. Even passing the value vers=2.0 to my Linux server, the mount wasn't working. So I had to enable on my Windows server smbv1 support. This article helped me: https://support.microsoft.com/en-us/help/2696547/how-to-detect-enable-and-disable-smbv1-smbv2-and-smbv3-in-windows-and