I have a bash script that uses sshfs to remotely mount an SSH directory.
Where should i place this script so it mounts with route privileges?
One consideration is i use VPN so the directory needs to be mounted after openvpn is started.
We recently had a serious loss of data because of SSHFS mounting. A user in our group mounted the entire home directory of our server (/home). This was so they could easily move between user folders to read/write data from other people involved in the same project. They then deleted several folders that were not in there home directory.
Now I know this is a bad idea and that there should be a dedicated "projects" folder where everyone collaborates and does their stuff. Such a folder/system exits but I can't make them use it. My question: Is there anyway to configure SSHFS such that the only thing the user's can mount is their home directory? Obviously this won't fix the problem since they can sym-link to other folders but I've got to start somewhere. Perhaps there's a better solution (one that doesn't involve me nagging users about proper form).
I have an Ubuntu 10.04 machine at home and apache setup on it (files are located in a Truecrypt volume). The reason for the web server being that I wanted to access my files wherever I'm at (i.e. hotel, work, hotspots, etc...). So far, it's worked out great for me seeing as a I can http download my files (or stream media files). However, I am often on a public hotspot and I know it's a matter of time before someone finds the webserver on my computer. I have the machine firewalled and password protected (via .htaccess), but either way I don't want people looking in on my computer.
The problem: I have used Truecrypt for a long time and completely trust using the program to encrypt/unencrypt a volume container to store my files. Usually, I would remote desktop into my computer and mount/unmount the volume when I needed it. However, after time it get's really annoying to do this. So I eventually figured out how to setup a bash script to automatically do this for me (which I put on the usb part of my phone). What I wanted to do was to be able send the bash script to my Ubuntu machine (via ftp from my phone) and have Ubuntu automatically run the script. Is this possible? What programs do I need on Ubuntu?
I was thinking about using something like cron, but that is for scheduled times. I don't really have a set time in which I need my files, it's pretty sporadic depending on how much I travel. Thus the need for being able to remotely mount the volume when I need it.
Summary: I need a way for Ubuntu to read a folder every minute or so to check for bash scripts to run. I want to be able to send the bash script via ftp from my phone, have Ubuntu run the script, then delete itself (so as to not store the password). I already know the script in which to mount the Truecrypt volume and how to send the file via ftp from my phone. It's really a matter of what program to use in Ubuntu to find and run the script.
I am struggling with getting an sshfs mount mounted on system boot. I have a script that mounts the sshfs for "userA". When userA runs the script all is well - user A can access the remote filesystem, root user can't see it as expected. The basic command is: sshfs userA@remote host:/home/userA /home/userA/mountdir -p 21212 -o password_stdin < passwordfile. I can prepend the sshfs command in the script with su - userA -c and when I run this script logged in as root all is well, userA has access and all is well. If I then put this script in /etc/init.d and reference it properly in the rc. directories the mount doesn't happen. If I prepend the sshfs command with sudo, same thing. Logged in as root I can run the script and UserA has access. Run the script in /etc/init.d during startup and the mount doesn't happen. Echoing text to a log file shows that the script is being executed but no mount happens.
View 4 Replies View RelatedI use this command to mount sshfs:sshfs -o idmap=user user@ip:/home/user/public_html ~/FolderThen I enter my password. I do this every time I start my computer
View 3 Replies View RelatedI've got a non-vital sshfs filesystem in my fstab (by non-vital I mean just files that I access on-demand).However, the fstab file seems to get read and things try to mount before the network has been brought up. I recently switched to Kubuntu 10.10 from Arch, and Arch didn't display this behaviour.I'm wondering if it's possible to make it mount the sshfs filesystem AFTER the network has been brought up?
View 2 Replies View RelatedI'm trying to set up sshfs on a desktop/fileserver so that I can mount its HDD's I use for storage on my laptops. My issue seems to be getting permission to write to the mounted drives. I just keep getting permission denied. here is what I think are some relevant tidbits.
mount point on server,
Code:
joe@joe-desktop:~$ ls -ld ~/sda2
drwxr-xr-x 2 root root 16384 1969-12-31 19:00 /home/joe/sda2
HDD I'm trying to mount
Code:
joe@joe-desktop:~$ ls -ld /dev/sda2
brwxrwxrwx 1 joe disk 8, 2 2010-10-17 15:22 /dev/sda2
Mount point on client
Code:
joe@joe-laptop:~$ ls -ld ~/sda2
drwxrwxrwx 2 joe joe 4096 2010-10-17 19:58 /home/joe/sda2
This is the command I'm trying to use to mount
Code:
sshfs joe@192.168.1.103:/home/joe/sda2 ~/sda2
I can chmod the mount on the server to 777 but as soon as I mount the HDD to the server all write permission just goes *poof* and its back to "drwxr-xr-x".Chown seems either, it just wont stick.
I have a problem with sshfs. I want to share a binary with some others computers, but i only want them to be able to execute (no read/write ). So, on my main server, I chown root:root bin & chmod 701 bin. That work nicely on main server, local users can execute bin w/o read/write ... But when I mount directory using sshfs, users cant exec/read/write ...
SSHFS version 2.2
FUSE library version: 2.8.4
fusermount version: 2.8.4
using FUSE kernel interface version 7.12
I'd like to add a user to my server that will only have access to a mount point over sshfs. Is there any way I can provide them this access without actually giving them permission to open a terminal on my server? I tried /bin/false and /sbin/nologin already, but /bin/false didn't allow the mount point to be made and /sbin/nologin prevented a login completely (also stopped the mount point from working).
View 6 Replies View RelatedI have a server with a /data/ directory, everything in the /data/ folder has "-rwxrw-rw- 1 root root" permissions.all works fine, multiple users are mounting this over a lan and everyone is able to modify files. However I would like to be able to access the /data/music/ directory from the internet.
Is it possible to configure sshfs to only accept logins from a user restricted to reading the /data/music directory, or would it be possible to tunnel nfs over ssh in such a way that everyone on the lan 192.168.0.xxx has unrestricted access to the data directory, but something coming from outside only has read access to the music directory. Although is one were tunneling nfs over ssh, the nfs mount request would appear to come from the server itself. The router is at 192.168.0.1 and the server is at 192.168.0.3.This Seems very much like what I want to do, however I'm having a bit of trouble getting this to work well with other users mounting with full rights over the lan too.
Having some troubles with sshfs + autofs. Roughly following this guide here: [URL] Here is /etc/auto.master:
[Code]....
But when I 'cd' to the dirs /mnt/sshfs/remote1 or /mnt/sshfs/remote2, I get this error:
Code:
bash: cd: remote1/: No such file or directory
I'm working on my first bash script. My script will do several things but right now I'm just trying to get the basic part of it down and working.I have a section that looks like this
Quote:
#/!bin/bash
SERVER=$1
if [ ! -d `ssh ${1} /somedir`];then
echo "Bad"
[code]...
The problem is that if you take that right now and run it, it will return back good in that it does exist. What I need it to do is pass back that it's bad because it doesn't exist (that is unless you actually do have that directory in your root).
I have a computer (C1) to which I can connect through the Internet (ssh, for instance) (it has a static ip and though it is sitting behind a router, the appropriate ports are all forwarded). I have another computer (C2) that doesn't have a fixed ip address and sits behind a router that I cannot fiddle with (so no port forwarding here). I would like to know if there is any way to connect from C2 to C1 such that a directory on C2 would be mounted on C1.
View 5 Replies View RelatedI finished installing Lucid (Ubuntu 10.04) on two desktop PCs formerly running Windows XP. Both installs went like a charm.Machine A has an external USB drive connected to it that I previously used for MS Windows backups. I can see the contents of the external drive from A's "file explorer" and navigate its directories. I would like to have machine B auto-mount A's external USB drive.A while back, I did something similar with a RedHat host that would automount a connection to an external USB drive shared by a MS Windows machine. My "/etc/fstab" was updated to mount a CIFS share.
I used CIFS since RedHat does not support NTFS, which is the format of USB drive. It worked perfectly.Can I do the same thing or is there a simple way? I am thinking that I may need to setup NFS server and client but I am not sure.Is there a how-to on NFS and Ubuntu?
I've got a problem about how to map or mount the whole developing environment remotely?
For example,There is a server A (CentOS) which I have got a normal user name and password but no root control.And my desktop B (Opensuse) which I've got the root authority.Normally, when I do some software developing, I use SSH to login to A. All the developing environment are installed on A, e.g. libraries, PATH, makefile, etc.However, there is a waste of my machine B, because all the test, debugging are performed on A.
I know that sshfs can map a path to local remotely. May I ask if there is a method to map or mount the whole working environment from A to B please?
I've got a problem about how to map or mount the whole developing environment remotely?
For example,
There is a server A (CentOS) which I have got a normal user name and password but no root control.
And my desktop B (Opensuse) which I've got the root authority.
Normally, when I do some software developing, I use SSH to login to A. All the developing environment are installed on A, e.g. libraries, PATH, makefile, etc.
However, there is a waste of my machine B, because all the test, debugging are performed on A.
I know that sshfs can map a path to local remotely. May I ask if there is a method to map or mount the whole working environment from A to B please?
I've run into a problem with my bash program.
Code:a mkdir -p /home/chris/Downloads/ARCHIVE/`date -I` find /home/chris/Downloads* -mtime +14 -exec "cp {} /home/chris/Downloads/ARCHIVE/`date -I` ;
It's meant to move old files from my Downloads folder into an archive file (later tar them). The directory exists, I've used $PWD and nautilus and ls to make sure it's there, yet for every file it gives
Code: find: `cp /home/chris/Download/foo.bar /home/chris/Downloads/ARCHIVE/2011-08-19': No such file or directory I tried to test it on a different destination, specifically my home folder. IT still gave me the error.IS it a bad syntax within find or something else? I'm running a 32-bit system with 11.04
I can set my EDITOR variable in my .bashrc to e.g. gedit, which is a nice graphical editor on ubuntu. But when I log in over PuTTY that editor will fail.What is the best way to detect if I'm logged in and can run a graphical editor, or not, and set EDITOR appropriately?
View 2 Replies View RelatedRecently I mounted a larger partition into my home directory since I was running out of space, Everything went smoothly, but it caused me to wonder about something I cant figure out. While playing with the mount unmount commands when I was copying everything over... before editing my fstab.
Is there a way to access the files that existed in a directory before you mount a partition to that directory? after mount the original files are gone.unmount and they are back, Where do they go?
I have Ubuntu Karmic. I chose to install with an encrypted home directory. Recently I got a warning that I only had 2GB of drive space left. This is mostly because of my videos. So I went and bought a new hard drive and partitioned it and made 1 ext4 partition and copied my videos all to the new hard drive. I added a line in my fstab to mount the new hard drive to ~/videos, but when I reboot the computer, there is a screen saying something like "error mounting /home/me/videos, press S to skip or something else to reboot". If I press S to skip, then when my system comes up there is a video directory but it's empty because my other hard drive didn't get mounted. I can run sudo mount /dev/sdb video/ and it will mount fine and I can see all my videos, so why can't fstab mount it? Does this have something to do with my encrypted home directory?
View 14 Replies View RelatedI have a requirement that seems to be unique in nature. I have multiple clients who are caged to their home directories. I would like to "share" a directory which exists above these chroots with all these caged users. I know this can be accomplished using mounts but my problem is, how can I mount a single directory to multiple mount points located in each users home dir? Can this be done in the fstab file?
View 9 Replies View RelatedI have a server with Fedora 13 with which I would like to get NFS working. I have looked up multiple howto's and tutorials, but I'm having a problem not addressed by any of them.Official how-to, another how-to, and another how-to.I have verified that nfs-utils, nfs-utils-lib, portmap, and system-config-nfs are installed and running. I have verified that I have, in fact, shared the directory that I want to share, and that the proper permissions are set.
I had to go through some gyrations to get the Belkin wireless N router to allow my server to have a static IP. However, I can ping the server from the nfs client (a toshiba satellite running mint 8), and vice versa. I have (for now) disabled firewalls on both computers. I think I have disabled SELinux on Fedora 13 (for now).When I attempt to connect to the server from the client, the output looks like this:Quote:
aragorn ~ # mount -v 192.168.2.101:/test /home/kelev/test/
mount: no type was given - I'll assume nfs because of the colon
mount.nfs: timeout set for Sat Dec 18 12:21:09 2010
[code]....
I'd like to mount the directory /var/www/mysite to the directory /home/daniel/mysite, but also have the user of the mounted files mapped from the original user (www-data) to my own user (daniel). So that the file /var/www/mysite/index.php who's user is www-data will appear in the mounted directory as /home/daniel/mysite/index.php and be owned by daniel - and alternatively, if I create a file /home/daniel/mysite/test.php with my own user, it will be created in the original directory under the user www-data Is it possible? If not, what alternatives do I have so I can use an IDE and still make sure all the files belong to the HTTP server's user?
View 2 Replies View RelatedDoes anyone know or recommend some software or a script to remotely power on a PC from standby to on, or even better from completely off?
I guess the completely OFF to ON is much more complicated - would probably require an extra piece of hardware(?)
We currently have a NFS shared Directory mounted as read-only on our server.This directory contains multiple sub-directories and files. It being read-only is a equirement. Now, we need a directory underneath to be read-write. Is there a graceful way to make that happen? Like a special mount option to use? Basically objective is: /u01 is mounted as read-only and has 3 directories: dir1, dir2, dir3 dir3 has 2 sub-directories- sub1,sub2/u01/dir3/sub2 needs to be read-write, while all other are read-only.
View 2 Replies View RelatedHow can i auto mount more than directory in the same directory ? i want to automount 2 home directories in the /home and still be able to enter the other home directories !
The problem that i've another account on the system with home directory joe when the user1 home directory auto mounted i become unable to enter joe home directory !
I'm not terribly new to Linux, but I am new to the forums, so hear me out! I am in the process of creating an electronic mapwall for our meteorology program, and have designed the computing system from scratch. I have two Linux Boxes, each with capabilities for 6 attached monitors...a total of 12 displays driven from two machines. My intention is to have one machine be the master...it has a touchpanel control. The inputs to the touchpanel will then trigger events for the both the master and the slave machine to display. Each of them has a specific IP address (DNS entry), and are not on a subnet.
Now...is there a way to remotely login to the slave machine and have it display on it's OWN monitors? The code is Java and which works on the master machine to animate directories of .gifs for each of the master's attached monitors. I will most likely have Java execute shell commands for the remote login (ssh), but I believe the answer lies somewhere in the X-configuration. Do I have the machines in an adverse configuration (creation of a subnet would be better)? Lots of questions...lots of desire...few answers!
I had netbeans installed and I removed it using wheris netbeans and then sudo rm whatever files it gave.I also remember doing sudo rm -r /usr/lib/netbeans6.9 I did this because i wanted to install netbeans via the repo. Now I installed netbeans suing
Code:
varun@varun:~$ sudo apt-get install netbeans
and when I try running netbeans i get...
Code:
varun@varun:~$ netbeans
/bin/bash: : No such file or directory
Have i changed permissions or something?
I've written a script which does various rsync commands to a Linux server where I backup all my files.My Ubuntu (Lucid) PC has 3 internal drives in it, the Ubuntu partations are all on one disk, then I have 2 further disks both have a single NTFS partition (leftover from my Windows days).When I want to mount one of the NTFS drives I just hit the drive name in Nautilus, handily they are both listed in 'Places'nted in:/media/FDrive/media/DDriveBack to my rsync script - it rsyncs files from both /media/FDrive and /media/DDrive. So at the moment I make sure both disks are mounted before running my rsync script, if they're not I just mount them by clicking on them in Nautilus before running the script.But now I want to run my rsync script using a cron job. So I've written a test in the script to check if they are mounted, like this:
Code:
cat /etc/mtab | grep /media/DDrive > /dev/null
RetVal=$?
[code]....
An executable fails to be found and run when called even though it is found and displayed by ls.I am on lucid LTS fully updated 32 bit.The file has permissions setup but any attempt to run it results in the error bash ./executablename No such file or directory
This situation also seems to be present in kcronsometimes my scripts dont run from it even thoughthe run now option does.This bug has been around since dapper it would be nice if someone could fix it.