Strange problem. We use a cf web server to access NAS. We have tested this issue to the nines and can't figure out the cause.
The problem: We get empty result sets doing a cfdirectory on the share to a known directory with the correct casing. However, we can ls from the cf server as the cf user and see everything without permission errors.
Tests we've tried:
Making a test file to be user the path we are testing is correct - fail.
Doing a directory listing from python - works.
Doing a CFFILE read and write from the offending web server to the directory in question - works.
Doing a CFDIRECTORY on a local directory - works.
Doing ls -la on the directory sudo'd to cfuser - works.
Doing ls -la as root on the directory - works.
Changing cf user permissions to root and retrying CFDIRECTORY - fail.
Changing mount to mount as root user and retrying CFDIRECTORY - fail.
chown-ing the files and the parent dir and retrying CFDIRECTORY - fail.
I can only think of a couple of things. First, make sure your case is correct since filesystem reads are case-sensitive in CF in Linux.
Secondly, I have not had much luck reading directly to SMB from CF. What has worked for me in the past is mounting a drive using SMB Fuse as a normal mount point and making sure the owner/group matches the CF user.
Related
So I have an interesting issue that I just can't figure out why I'm getting this and what to do.
So basically I store all my development projects on my Synology NAS for local access between my various devices. There has never been a problem with this until I started playing around with Elixir and more importantly Phoenix. The issue I am getting is when running mix phx.server. I get the following
[warn] Phoenix is unable to create symlinks. Phoenix' code reloader will run considerably faster if symlinks are allowed. On Windows, the lack of symlinks may even cause empty assets to be served. Luckily, you can address this issue by starting your Windows terminal at least once with "Run as Administrator" and then running your Phoenix application.
[info] Running DiscussWeb.Endpoint with cowboy 2.7.0 at 0.0.0.0:4000 (http)
[error] Could not start node watcher because script "z:/elHP/assets/node_modules/webpack/bin/webpack.js" does not exist. Your Phoenix application is still running, however assets won't be compiled. You may fix this by running "cd assets && npm install".
[info] Access DiscussWeb.Endpoint at http://localhost:4000
So I tried as it stated and ran it in CMD as admin but to no avail. After some further inspection I tried to create the symlinks manually but every time I tried I would get a Access is denied. error (yes this is elevated CMD).
c:\> mklink "z:\elHP\deps\phoenix" "z:\elHP\assets\node_modules\phoenix"
Access is denied.
So I believe it is something to do with the fact that the symlinks are trying to be created on the NAS because if I move the project and host it locally it will work. Now I know what you're thinking. Yes, I could just store them locally on my PC but I like to have them available between PCs without having to transfer files or rely on git etc. (i.e. offline access), not to mention that the NAS has a full backup routine.
What I have tried:
Setting guest read write access on the SMB share
Adding to /etc/samba/smb.conf on my Synology NAS:
[global]
unix extensions = no
[share]
follow symlinks = yes
wide links = yes
Extra logging on SMB to see what is happening when I try it (nothing extra logged)
Creating a symbolic link from my MAC (works)
Setting all of fsutil behavior query SymlinkEvaluation to enabled
At the moment I am stuck and unsure of what to try next, or even if it is possible. Considering just using NFS instead but will I face the same issues with SMB?
P.S I faced a similar issue with Python venvs a while ago, just a straight-up Access is denied. error and just gave up and moved just the venv locally and kept the bulk of the code on the NAS. (This actually ended up beingthe best solution for that because the environments of each device on my network clashed etc.)
Any ideas are greatly appreciated.
I went to vagrant up this morning and got this error.
VirtualMachine: Cannot create bind mount from '/mnt/vagrant-vflsf_7p872mq5085x' to '/var/www': Destination path '/var/www' is already a bindfs mount
Made no updates to my vagrant configuration (1.9.1) or my operating system (MacOS 10.12.2).
I've had the same problem today, which turned to be due to some old invalid entries in my /etc/exports file on my host machine from destroyed vagrant boxes.
Open your /etc/exports file and remove any invalid entries (and/or edit the paths for existing entries).
I needed to also do a vagrant reload.
This can happen if two filesystems have similar names. The fix is to order them from least unique to more unique in whatever file you store your bind mountpoints in.
For instance, say you mount /abc/this_dir and /abc/this_dir_too
This will NOT work
/abc/this_dir_too
/abc/this_dir
This WILL work
/abc/this_dir
/abc/this_dir_too
The problem is that vagrant issues a really stupid command to find if a mount has already occurred:
mount | grep '^bindfs' | grep #{directory.shellescape}"
It's pretty obvious what will happen if one FS is a subset of the other.
I have an embedded system running linux and I have busybox installed for running different services. One of the services is ftpd, which I start like this from the etc/init.d/rcS file:
tcpsvd 0.0.0.0 21 ftpd -w -v /mnt/flash&
I have two users in the system, root and a regular user. Root user can easily connect to the FTP server, but when I try to login with the regular user's credentials, I get this error:
ftpd[678]: can't change root directory to '/mnt/flash': Operation not permitted
Now, I thought that it must be a directory permission problem, and I started with changing permissions on the /mnt/flash directory first, but after this didn't work, I ended up having all my files in file-system including / to be owned by this regular user and have drwxrwxrwt permissions (just to find the cause of the problem). But I still get this error.
I have also tried to start ftpd with different root folders, including /.
I also couldn't find any config files for the ftpd in my file-system, maybe I need to create one manually? If so, which one and how to tell ftpd to allow user logins?
I'd really appreciate any help. Thanks.
Well, after trying some random things, I discovered that user login works fine if there is no directory changing involved at all:
tcpsvd 0.0.0.0 21 ftpd -w -v &
I suspect that chroot is called otherwise and the user does not have a right to do that.
I am new to Vagrant and get the following error on vagrant up or vagrant ssh:
The private key to connect to this box via SSH has invalid permissions
set on it. The permissions of the private key should be set to 0600, otherwise SSH will
ignore the key. Vagrant tried to do this automatically for you but failed. Please set the
permissions on the following file to 0600 and then try running this command again:
[...]/.vagrant/machines/default/virtualbox/private_key
I have run:
$ sudo chmod 666 [...]/.vagrant/machines/default/virtualbox/private_key
I also tried (600, 777) but still get the same error.
Please can someone tell me what is wrong and how to fix it?
I just had this issue, and I worked around it moving the private_key file to another place, changing its permission, and then creating a symbolic link at the original place.
So,
$ mv [...]/.vagrant/machines/default/virtualbox/private_key /some/path/where/you/can/change/permissions
$ ln -s /some/path/where/you/can/change/permissions [...]/.vagrant/machines/default/virtualbox/private_key
If you're using the Windows Subsystem for Linux (WSL), this error can occur when you're trying to vagrant up in a directory that is outside the user's home directory.
From the Vagrant docs:
If a Vagrant project directory is not within the user's home directory on the Windows system, certain actions that include permission checks may fail (like vagrant ssh). When accessing Vagrant projects outside the WSL Vagrant will skip these permission checks when the project path is within the path defined in the VAGRANT_WSL_WINDOWS_ACCESS_USER_HOME_PATH environment variable.
Changing the VAGRANT_WSL_WINDOWS_ACCESS_USER_HOME_PATH to the current working directory (or a directory above it) can fix this. For example, if your project is in /mnt/c/www, then set the environment variable accordingly:
export VAGRANT_WSL_WINDOWS_ACCESS_USER_HOME_PATH="/mnt/c/www"
I got the same error now. The problem happened because i was trying to do vagrant up in an NTFS partition, just like the error message tell me.
So i created an directory link in my ext4 partition and an simbloc link in my NTFS to solve this. Works Fine now!
Thanks!
I had this same problem and turns out chmod seems to be working fine but is not actually changing permissions, my files where at an NTFS partition, try changing them to an ext4 or similar.
Got this error using otto (which layers on vagrant)
It is def filesystem related, have a fat partition to allow use with windows (used to, no longer). When the permissions couldn't be set on the partition I just copied the whole directory over to my user directory (as I always should have).
Was using git so I just reset to head to get back to my starting place... re-ran:
otto compile
otto dev
up and running now.
Based on a recommendation on stack overflow, i am giving yii framework a spin.
It appears that I am only able to gain shell access to my web server if I log in as root.
I am running through the tutorial, and am having a number of problems typical command to create a web app-
/usr/local/bin/php /home/testuser/public_html/yii/framework/yiic webapp /home/testuser/public_html/demo
Do I have to type /usr/local/bin/php EVERY TIME???
Do I have to type /home/testuser/public_html/yii/framework/yiic in full every time???
Do I have to yype /home/testuser/public_html/demo in full?
As I am logged in as ROOT, I have to chown every file / folder created by root to make the file viewable by testuser!!!
I create a new controller using yii. controller message helloWorld, yii does some funky stuff, but I can't do ANYTHING with the files without chown EVERYTHING!!!
I'm not too confident with SSH, and I know it's all a learning curve, but... having to change file permissions EVERY time, and having to type full directories every time is just a pain.
I know that there are some shortcuts that I will probably learn with time, but if anybody could give me some pointers to make my life about 1M times easier (considering my main objectives at present are to learn yii framework) linux/terminal/shell commands are currently secondary to my requirements.
Here's what I'd do, in this order:
1 - When you login as root issue the command su - testuser. This changes your effective user id and groupid to that of testuser, meaning, you now are operating as that user account. Your current directory should now be the testuser's home directory. To get back to root user just type exit.
2 - Try running php -v without the fullpath. If it prints out a version number then you don't have to use the fullpath to run php. If you get an error then you can add /usr/local/bin to your PATH environment variable in your .profile file which is present in testuser's home directory. If the .profile doesn't exist then you can create it.
3 - Add /home/testuser/public_html/yii/framework to your PATH in testuser's .profile.
After doing all that then the next time you login as root you would be able to do the following:
su - testuser
cd public_html
php yiic webapp demo
And you wont have to worry about permissions since you are creating directories/files as testuser, not root.
See also: StackOverflow question about modifying your PATH