For mount points, I like to create an "offline.txt" file in there so that when the mount is successful, you do NOT see the text file. But if you do, you know the mount is "offline" hehehe.
LHammonds
For mount points, I like to create an "offline.txt" file in there so that when the mount is successful, you do NOT see the text file. But if you do, you know the mount is "offline" hehehe.
LHammonds
I wrote a too-long reply then, just prior to posting, closed the browser. Ouch.
I'll try to be more concise, for my personal sanity.
Check real storage mounts on any system using
Most of that command is options to prevent non-storage stuff from being displayed. We only care about real storage.Code:df -hT -x squashfs -x tmpfs -x devtmpfs
Check nfs exports on the nfs-server using.
If you change anything in the /etc/exports file, you need to restart the nfs-server, do that withCode:showmount -e
We all forget to do that sometimes. To get the current status, useCode:sudo service nfs-server restart
Log files and errors can be checked in many different ways, but journalctl -xe is usually the easiest.Code:sudo service nfs-server status
If you change an /etc/fstab line, you'll need to un-mount the file system, then mount it again. It works for all storage, local or NFS.
NFS will not ask you for credentials to mount anything. It isn't an end-user mount system. It is computer-to-computer.Code:sudo umount /Vol10 sudo mount -a
LHammonds has a good tip with the empty file so you'll know if the mount happened. It is easy to programmatically check in a script as well.
SeijiSensei has good ideas too. The differences are from our different experiences and different work environments.
All of us have lots and lots of experience and long histories here trying to help people. None of us has all the answers. We all make mistakes. The first time we each setup NFS, you can bet we struggled a little too.
With Unix, there is almost always 50-500 different ways to accomplish the same thing. The days of 1-right-answer seldom exist on Unix systems.
On the NFS server, you'll probably want to make your userid the owner of /Vol10 and any subdirectories after it is mounted there. This will carry over to NFS clients with that userid as the owner. My NFS tips above tried to explain the relationship between the userids on each system. If there are issues - like different names are seen on different systems, then you'll want to post the 'id' command output for each userid on each client system. With only 1 client system, there won't be issues.
Try to mount the share manually with
The "-v" switch means "verbose." You may see where the problem lies.Code:sudo mount -v bigbin-xubuntu:/Vol10 /Vol10
Sorry I missed the absent colon in my earlier post.
Last edited by SeijiSensei; November 30th, 2020 at 09:32 PM.
If you ask for help, do not abandon your request. Please have the courtesy to check for responses and thank the people who helped you.
Blog · Linode System Administration Guides · Android Apps for Ubuntu Users
A hearty thank you to all.
Just prior to getting these posts I tried an experiment turning both machines off and then only turning the client on. The file Vol10 was still there in the files list. But now hen I try to restart the server it goes into emergency mode. I typed journalctl -xb, but there was pages of output. I think I need to get to the fstab file to comment out my last entry. I can get to the root command prompt and cd into ect - root@bigbin-xubuntu:/etc# but attempting to edit fstab results in
"unable to init server: could not connect; connection refused".
Unfortunately I have to leave this for now. If all goes well I will have a short spell tomorrow afternoon and a whole day at it on Friday. There is a journal file I kept on this machine that I should have saved elsewhere but didn't, so I hope I'm going to be able to get back in.
Couple more things.
First, I don't understand "offline text file". I would presume write a text file in the shared directory so it's obvious if it's loaded, but then you'd see it if its mounted and not if it isn't. This sounds a good idea, but can I have a bit more detail please.
Second, if i cant't get back into the server I'll have to start again. Not such a big deal, this is the sandbox. Things that have happened mean that this machine is destined te become a full time server, rather than a workstation with server modules added. Would now be a good time to do it. If I install Ubuntu server will the rest of the setup be very different. I intend to do this eventually, but it will only be now if I have to re-install the software anyway.
Ubuntu Server doesn't have a GUI. That makes it more stable. GPU drivers have always sucked a bit on Linux - more so recently.
If you have another workstation and can ssh into the server system using it, which is how Unix admins have worked for 40+ yrs, great.
Assuming /Vol10 is your mount point and the mount point NOT mounted, run these commands:
When you type "ls /Vol10" and see "/Vol10/offline.txt" then you know it is not mounted.Code:sudo echo "This file is used to tell if the mount is active or not" > /Vol10/offline.txt sudo chown root:root /Vol10/offline.txt sudo chmod 0444 /Vol10/offline.txt
After you issue the mount command, do the list again and if you no longer see the offline text file, then the mount worked.
If at any point after the volume is mounted that you ever see the offline file, that will be a clear indicator that it is dismounted (besides the fact that your files are gone...but at least this will let you know it is dismounted rather than deleted).
It seems you plan to have it mounted all the time so something like this will help let you know if it stops working.
Most of the mounts I have used were always temporary so I wrote a script to toggle it on or off. When run, the script would check if the offline file is there...if so, I mounted the share. If not, I dismounted the share...which effectively acted like a toggle without me specifying if it should be mounted or not.
LHammonds
Offline text is going to be usefull. Thanks.
The machine that is now my server used to be the main Linux workstation. At the beginning of this I bought a decent second hand motherboard and I7 processor. I just havent had time to put it in it's case. With a bit of luck, and assuming I can work out the f-panel connections without a manual, I should have a new workstation by the end of tomorrow.
It sounds as if there are other questions to answer if I go fully into the server route, however, so in the interest of "one thing at once" I'll hold off that for now.
How will I go about getting into bigbin-xubuntu without re-installing everything. It's currently stuck at the root command prompt in emergency mode.
If you have root, then you own the box and can do anything. Undo whatever you did to break it.
If you don't know what you did to break it, but think it is related to some userid settings, just move those settings somewhere else.
logout, login again. Simple.Code:mv ~/.config ~/.config-old
If you don't know which settings you modified to break it, just create a new userid with a new HOME directory, being careful to get the groups just like they were for the other userid that you broke, especially the sudo and adm groups. There are two different commands for creating a new userid. I never remember which does everything (creating a HOME, copying in skel files, etc.) and which is low-level useradd or adduser. I always have to check the manpage. These commands are ubuntu only, I think. Other distros have slightly different versions, so read the manpage ... or use the GUI admin tool, if you have one. Double-check the groups.
If you don't know how to accomplish these things - http://linuxcommand.org/tlcl.php is a good beginner-level book.
Last edited by TheFu; December 4th, 2020 at 05:57 PM.
Bookmarks