I'm following the Scripts with Support Files answer from here https://stackoverflow.com/a/46479538/4771016 which works great but running into a problem during the update of my script.
If not found, my script creates an .env file for the users to pass some variables in the same directory as the .sh file lives: /home/linuxbrew/.linuxbrew/Cellar/myscript/1.0.2/libexec/.env the problem is that upon releasing a new version, the .env file won't be in the new directory i.e. /home/linuxbrew/.linuxbrew/Cellar/myscript/1.0.3/libexec/ and thus will be recreated losing the modifications.
Any ideas for keeping that .env file during updates or an acceptable design pattern for my use case? I was thinking about keeping the .env file outside that directory somewhere, but I don't know the Homebrew directory structure well enough to store it in the right place.
Related
I'm using dask_jobqueue to establish a SLURMCluster. I'm trying to pass python files in the parent directory to the workers. I tried different ways including sys.path.append, setting PYTHONPATH in my .bashrc file, and setting PYTHONPATH in env_extra (it seems DASK version can't recognize job_script_prologue), but they all fail. The script works only when the imported Python files are in the same directory as the script. I just wonder if this is really the case (which means I should re-write my code to keep all files in the same directory), or I'm missing something here?
Also, I tried to use with performance_report to record some statistics in an HTML file, but it seems it does not work with dask_jobqueue, right?
I am trying to update a module to a newer version. In the past I have manually uploaded each file carefully into the new directory and overwritten older files using FTP. However I wanted to use SSH to try and do this more easily and without any file permission problems.
I have:
Uploaded the .tgz file to the root folder (/http) on the server
Logged into the server via SSH
Changed the directory to the correct directory
Run the following command: tar -zxvf fishpig_splash.tgz
In the command line I was then given a list of all the files that had been extracted. However if I use FTP to go to any of these files I can see that they are still the older version and have not been overwritten.
I was expecting that the files would extract into the correct directories and overwrite any that already existed. I have tested the extraction by creating a temporary directory and extracting into that and everything worked fine.
Is there another part to this script I need to use to overwrite the files?
Thanks
Glynn
Sorry this was just me being stupid! When extracting the tar file there was a subfolder within it for the extension, I completely missed it. I just went down a level in the file and zipped up the contents only then extracted them at the root and everything worked fine. Thanks for the help though!
My .csv recreates on every release, and, as I understand, to keep its data unchanged between deploys I need to put it in /shared directory and simlink to it from my deploy.rb.
Is this the right route? (I have this question because I don't seem to find much info on how to do this with respect to, eg, databases, for some reason. /shared directory is mostly used for .conf files and paperclip-like directories).
When using capistrano, your application code will be "uploaded" to some directory on the server. Capistrano uses this structure:
/path_to_folder:
current - symlink to the directory with the current release
releases - contains all kept releases
shared - files that should persist between releases
So to your question - copy the .csv file somewhere into "shared" directory and then in the config/deploy.rb add this:
namespace :deploy do
task :create_symlinks do
run "ln -s #{shared_path}/something.csv #{latest_release}/db/something.csv"
end
end
after 'deploy:update_code', 'deploy:create_symlinks'
Replace "something" with the file name that you copied. You can also put the csv file into some directory under "shared" if you want to, I'd use "db" in this case. If you do so, don't forget to update path in the symlink.
Error message :
"svn: Can't open file '/Users/username/Projects/myproject/trunk/project/.svn/text-base/filetoupdate.h.svn-base': No such file or directory"
Question:
I have an issue I've replaced a file in a project (in Xcode) with a new file (For reference and if this makes a difference, the new file has the same name as the one I deleted previously).
Now when I try to commit my changes in Xcode I get the error message detailed above and am unable to commit the changes (i.e. adding the new file).
In the file system view (in Xcode on the left hand side of the screen) the file has an R next to it (indicating Replaced in the repository).
Does anyone know how to fix it so I can commit the files?
Thanks
There is a bug or limitation in Subversion when using case-insensitive filesystems:
https://superuser.com/questions/303348/cant-checkout-in-subversion
This bug normally shows up when checking out a repository that contains two files whose names differ only in case. Of course, these cannot exist at the same time in the same directory on a case-insensitive filesystem. SVN could give a much more helpful error message, but it can't really solve the problem.
Your issue is a bit different because I assume the file filetoupdate.h (with the old case) no longer exists in your filesystem. So it's not a case conflict in the working directory. But I guess that SVN is trying to create the file in .svn/text-base with the new case, while the old one still exists, and that is failing (for the same reason).
You could try deleting the file from Subversion first, keeping the local copy (untested). The new copy must be removed from SVN control for the commit to succeed:
svn rm --keep-local --force FileToUpdate.h
And the old copy must be removed as well, to allow us to add the new copy later:
svn rm --keep-local filetoupdate.h
Commit this change:
svn commit
Now hopefully you can add the new file to version control:
svn add FileToUpdate.h
If that doesn't work, you might need to blow away the whole checkout and start again with a fresh one.
Are you on a Mac or Windows? Those have case-insensitive filesystems which causes the above problem when
a file currently exists with the same name but with different cases.
To fix , checkout out the tree on a Linux machine, then "svn rm" one of the files.
Maybe your local version has permission issues. Check if your user have the permissions to write for the .svn directories.
good luck
It looks like something got confused somewhere. To fix, I simply copied the offending files, saved them under a new name. Removed the originals from the project and the added the copied (renamed) version of the file to the project.
It seems to be that SVN doesn't like it if you add and remove a file with the same name. I tried cleaning the SVN through terminal, but it had no affect on this issue. But changing the name did work for me.
I am attempting to mirror a directory on a remote server using rsync. However, I would like a copy of all newly created files to be stored in a separate directory on the local machine.
For example, if a new file is added on the remote server, I would like it to mirror regularly (for example, to ~/mirror), but save an additional copy of only the new file in another folder, (for example, ~/staging). To be clear, only the new files should appear in staging.
My first approach was to allow rsync to update the timestamps, and then use that to make a copy. However, I would now like to preserve timestamps.
Can anyone provide ideas on a simple approach? I am open to use of additional utilities other than rsync.
You might consider making hardlinks in the extra directory.
ln --force --target-directory=~/staging ~/mirror/*
Edit:
If this is a Linux system, incron will trigger on inotify events and would allow you to make copies of files as they are added to a directory you specify.