Client error 'mounting ... failed, reason given by server: No such file or directory'

[andrewklajman]

Apologies in advance for the amateurish question. I know that there must be something that I have done wrong but I am unsure what. If someone can help it would be appreciated.

On my docker server (192.168.1.39) I am running the following.

docker run \
-v /home/andrew/docker.uat/nfs:/container/path/foo \
-e NFS_EXPORT_0='/container/path/foo *(ro,no_subtree_check)'  \
-e NFS_DISABLE_VERSION_3=1 \
-e NFS_LOG_LEVEL=DEBUG \
--cap-add SYS_ADMIN \
--cap-add SYS_MODULE \
-p 2049:2049 \
erichough/nfs-server

The output when I run the above is the following.

==================================================================
      SETTING UP ...
==================================================================
----> log level set to DEBUG
----> will use 8 rpc.nfsd server thread(s) (1 thread per CPU)
----> building /etc/exports from environment variables
----> collected 1 valid export(s) from NFS_EXPORT_* environment variables
----> kernel module nfs is loaded
----> kernel module nfsd is loaded
----> setup complete

==================================================================
      STARTING SERVICES ...
==================================================================
----> mounting rpc_pipefs filesystem onto /var/lib/nfs/rpc_pipefs
mount: mount('rpc_pipefs','/var/lib/nfs/rpc_pipefs','rpc_pipefs',0x00008000,'(null)'):0
----> mounting nfsd filesystem onto /proc/fs/nfsd
mount: mount('nfsd','/proc/fs/nfsd','nfsd',0x00008000,'(null)'):0
----> starting rpcbind
----> starting exportfs
exporting *:/container/path/foo
----> starting rpc.mountd on port 32767
----> starting rpc.nfsd on port 2049 with 8 server thread(s)
rpc.nfsd: knfsd is currently down
rpc.nfsd: Writing version string to kernel: -2 -3 +4 +4.1 +4.2
rpc.nfsd: Created AF_INET TCP socket.
rpc.nfsd: Created AF_INET UDP socket.
rpc.nfsd: Created AF_INET6 TCP socket.
rpc.nfsd: Created AF_INET6 UDP socket.
----> terminating rpcbind
----> all services started normally

==================================================================
      SERVER STARTUP COMPLETE
==================================================================
----> list of enabled NFS protocol versions: 4.2, 4.1, 4
----> list of container exports:
---->   /container/path/foo     *(ro,sync,wdelay,hide,nocrossmnt,secure,root_squash,no_all_squash,no_subtree_check,secure_locks,acl,no_pnfs,anonuid=65534,anongid=65534,sec=sys,ro,secure,root_squash,no_all_squash)
----> list of container ports that should be exposed: 2049 (TCP)

==================================================================
      READY AND WAITING FOR NFS CLIENT CONNECTIONS
==================================================================

But when I try to mount the folder from the client machine (192.168.1.189) I get the following.

showmount -e 192.168.1.39
clnt_create: RPC: Program not registered

mount -vvv -t nfs -o vers=4,port2049 192.168.1.39:/container/path/foo /mnt/test/
mount.nfs: timeout set for Sun Nov 8 09:22:36 2020
mount.nfs: trying text-based options'port=2049,vers=4.2,addr=192.168.1.39,clientaddr=192.168.1.189'
mount.nfs: mount(2): No such file or directory
mount.nfs: mounting 192.168.1.39:/container/path/foo failed, reason given by server: No such file or directory

I feel like it has something to do with the RPC (but I only have a basic understanding of it). However when I set up a native installation (not docker) of the nfs-server package on my server then my client has no problem connecting to it.

I did log into the container and confirmed that the path /container/path/foo exists. While in the container I also confirmed that the /etc/exports was there and correct.

So I am unsure what the problem is or how to follow up any more (I have spent a day on Google with this problem and have not gotten anywhere). I have tried several variations of the above.

This is only for a home server setup but if anyone can suggest how I would appreciate it.

Thankyou

[andrewklajman]

Hi All, I was able to get this to work but i do not understand why. If I run the following on my client then the share mounts fine. mount 192.168.1.39:/ /mnt/test

So then I was wondering how to share multiple folders. I tested with the following command

docker run \
-v /home/andrew/docker.uat/nfs1:/nfs1 \
-v /home/andrew/docker.uat/nfs2:/nfs2 \
-e NFS_EXPORT_0='/nfs1 *(ro,no_subtree_check,fsid=0)'  \
-e NFS_EXPORT_1='/nfs2 *(ro,no_subtree_check,fsid=0)'  \
-e NFS_DISABLE_VERSION_3=1 \
-e NFS_LOG_LEVEL=DEBUG \
--cap-add SYS_ADMIN \
--cap-add SYS_MODULE \
-p 2049:2049 \
erichough/nfs-server

The above generated the following output

==================================================================
      SETTING UP ...
==================================================================
----> log level set to DEBUG
----> will use 8 rpc.nfsd server thread(s) (1 thread per CPU)
----> building /etc/exports from environment variables
----> collected 2 valid export(s) from NFS_EXPORT_* environment variables
----> kernel module nfs is loaded
----> kernel module nfsd is loaded
----> setup complete

==================================================================
      STARTING SERVICES ...
==================================================================
----> mounting rpc_pipefs filesystem onto /var/lib/nfs/rpc_pipefs
mount: mount('rpc_pipefs','/var/lib/nfs/rpc_pipefs','rpc_pipefs',0x00008000,'(null)'):0
----> mounting nfsd filesystem onto /proc/fs/nfsd
mount: mount('nfsd','/proc/fs/nfsd','nfsd',0x00008000,'(null)'):0
----> starting rpcbind
----> starting exportfs
exporting *:/nfs2
exporting *:/nfs1
----> starting rpc.mountd on port 32767
----> starting rpc.nfsd on port 2049 with 8 server thread(s)
rpc.nfsd: knfsd is currently down
rpc.nfsd: Writing version string to kernel: -2 -3 +4 +4.1 +4.2
rpc.nfsd: Created AF_INET TCP socket.
rpc.nfsd: Created AF_INET UDP socket.
rpc.nfsd: Created AF_INET6 TCP socket.
rpc.nfsd: Created AF_INET6 UDP socket.
----> terminating rpcbind
----> all services started normally

==================================================================
      SERVER STARTUP COMPLETE
==================================================================
----> list of enabled NFS protocol versions: 4.2, 4.1, 4
----> list of container exports:
---->   /nfs2   *(ro,sync,wdelay,hide,nocrossmnt,secure,root_squash,no_all_squash,no_subtree_check,secure_locks,acl,no_pnfs,fsid=0,anonuid=65534,anongid=65534,sec=sys,ro,secure,root_squash,no_all_squash)
---->   /nfs1   *(ro,sync,wdelay,hide,nocrossmnt,secure,root_squash,no_all_squash,no_subtree_check,secure_locks,acl,no_pnfs,fsid=0,anonuid=65534,anongid=65534,sec=sys,ro,secure,root_squash,no_all_squash)
----> list of container ports that should be exposed: 2049 (TCP)

==================================================================
      READY AND WAITING FOR NFS CLIENT CONNECTIONS
==================================================================

Now when I run mount 192.168.1.39:/ /mnt/test it mounts /nfs1. But when I run either mount 192.168.1.39:/nfs1 /mnt/test or mount 192.168.1.39:/nfs2 /mnt/test I get the same as above.

mount.nfs: timeout set for Sun Nov 8 09:22:36 2020
mount.nfs: trying text-based options'port=2049,vers=4.2,addr=192.168.1.39,clientaddr=192.168.1.189'
mount.nfs: mount(2): No such file or directory
mount.nfs: mounting 192.168.1.39:/container/path/foo failed, reason given by server: No such file or directory

There really is only one directory that I need to share so I should be fine with the above. But still I dont understand why this is not working.

Can anyone offer some insight?

Thanks

[jjwong0915]

@andrewklajman I encountered the same problem and found the reason is that nfs4 does not use an auxiliary protocol which allows multiple export entry directly anymore, instead it uses a virtual filesystem which has a root directory containing each export entry as a sub-directory. As a result, exporting multiple folders can be achieved by bind-mounting them into a single parent folder first.

---

For example, if you want to share /home/user1/.ssh and /home/user1/.local/share:

mkdir /tmp/nfs /tmp/nfs/ssh_folder /tmp/nfs/share_folder
mount --bind /home/user1/.ssh /tmp/nfs/ssh_folder
mount --bind /home/user1/.local/share /tmp/nfs/share_folder

Then, create the nfs-server container with following exports:

docker run --rm --privileged -v /tmp/nfs:/nfs -e NFS_DISABLE_VERSION_3=1 -e NFS_EXPORT_0='/nfs *(fsid=0)' -e NFS_EXPORT_1='/nfs/ssh_folder *(rw)' -e NFS_EXPORT_2='/nfs/share_folder *(ro)' erichough/nfs-server:2.2.1

After the server is started, you can now nfs-mount any of the folders:

mkdir /mnt/ssh_folder
mount <container_ip>:/ssh_folder /mnt/ssh_folder
mkdir /mnt/share_folder
mount <container_ip>:/share_folder /mnt/share_folder

---

I hope this will help.

[andrewklajman]

@jjwong0915 , Yes you were absolutely correct.

I spent a few hours looking for the solution myself Im just curious for future reference how you came across the solution?

Thanks again.

[jjwong0915]

@andrewklajman The solution comes from this link, which is found in the top result of searching keyword nfs4 multiple exports.

[sergiochoo]

I tried to set it up with nfs4 protocol, but no luck. So I masked rpcbind protocol with command "sudo systemctl mask rpcbind" then rebooted the server. This action resolved my issue. My docker-compose.yml

version: '3.8'
services:
  nfs:
    image: erichough/nfs-server:2.2.1
    container_name: nfs
    volumes:
      - '/nfs/data:/nfs'
      - '/nfs/exports:/etc/exports:ro'
      - '/lib/modules:/lib/modules:ro'
    environment:
      - NFS_VERSION=3
      - NFS_LOG_LEVEL=DEBUG
    cap_add:
      - SYS_ADMIN
      - SYS_MODULE
    ports:
      - 2049:2049
      - 2049:2049/udp
      - 111:111
      - 111:111/udp
      - 32765:32765
      - 32765:32765/udp
      - 32767:32767
      - 32767:32767/udp

[wood-j]

same issue @2021

[lhzw]

NFS V4 cannot work, always reports:

mount -v 172.22.1.3:/ /mnt
mount.nfs: timeout set for Mon Aug 29 11:24:57 2022
mount.nfs: trying text-based options 'vers=4.2,addr=172.22.1.3,clientaddr=172.22.1.2'
mount.nfs: mount(2): No such file or directory
mount.nfs: trying text-based options 'addr=172.22.1.3'
mount.nfs: prog 100003, trying vers=3, prot=6
mount.nfs: portmap query retrying: RPC: Unable to receive
mount.nfs: prog 100003, trying vers=3, prot=17
mount.nfs: portmap query failed: RPC: Unable to receive - Connection refused
mount.nfs: trying text-based options 'vers=4.2,addr=172.22.1.3,clientaddr=172.22.1.2'

mount -v 172.22.1.3:/data /mnt
mount.nfs: timeout set for Mon Aug 29 11:15:00 2022
mount.nfs: trying text-based options 'vers=4.2,addr=172.22.1.3,clientaddr=172.22.1.2'
mount.nfs: mount(2): No such file or directory
mount.nfs: trying text-based options 'addr=172.22.1.3'
mount.nfs: prog 100003, trying vers=3, prot=6
mount.nfs: portmap query retrying: RPC: Unable to receive
mount.nfs: prog 100003, trying vers=3, prot=17
mount.nfs: portmap query failed: RPC: Unable to receive - Connection refused
mount.nfs: trying text-based options 'vers=4.2,addr=172.22.1.3,clientaddr=172.22.1.2'

Maybe we should update the document about disable V3, it's so time-wasting, thank you.

[whatwewant]

NFSv4 is virtual filesystem, it must have only one root path with fsid=0 so just => -e NFS_EXPORT_0='/container/path/foo *(rw,no_subtree_check,fsid=0)' is ok (/container/path/foo is the root of NFSv4) then client command:

$ mount -t nfs <SERVER>:/ /mnt/nfs

instead of

mount -t nfs <SERVER>:/container/path/foo /mnt/nfs

and Another easy way to use NFSv4 Server in Docker:

$ curl -fsSL https://raw.githubusercontent.com/zmicro-design/service-nfs/master/install -o /tmp/install.sh && bash /tmp/install.sh

OR

$ curl -fSL https://zmicro.zcorky.com/ | bash
$ zmicro service start nfs

then Client

$ mkdir -p /mnt/nfs
$ mount -t nfs <SERVER>:/ /mnt/nfs