Closed BoBRoID closed 5 years ago
Hi @BoBRoID
Can you please tell me:
This may help a lot to identify a bug and fix it.
Thanks
Hi @gael-ian!
My VM's provider is libvirt, and OS are centos/6
and centos/7
.
I'm not very good at vagrant configuration, so I took yii2 default configuration, and just put vagrant-bindfs to the required modules list:
required_plugins = %w( vagrant-hostmanager vagrant-vbguest vagrant-bindfs )
required_plugins.each do |plugin|
exec "vagrant plugin install #{plugin}" unless Vagrant.has_plugin? plugin
end
I always uses:
cd /path/to/project sudo vagrant up --provider=libvirt
and machine started without any problems. But today I've got error message which is described in the first topic. After few hours I decided to install older bindfs version:
sudo vagrant plugin uninstall vagrant-bindfs sudo vagrant plugin install vagrant-bindfs --plugin-version=1.1
and then my machine started correctly, without any errors.
I hope this information was helpful.
Hi @BoBRoID
By default using CentOS, no packages repository containing bindfs is activated so it will always be installed from sources (See #83). Your problem was this installation failed because the install script could not download the correct tarball or build the sources it contains.
At some time between 1.12.0 and 1.13.0, the release tools used for bindfs has changed. Before 1.13.0, the tarball name will omit the patch release number if it was 0. Since 1.13.0, the full version number is always used.
I added a few method today to the plugin to ensure the proper tarball name will always be used to download bindfs sources and build them. This should fix your problem.
As there is no big differences between vagrant-bindfs 1.1.0 and 1.1.2, you can just use your installed plugin as is or you can upgrade to the latest version if you want.
Sorry for the inconvenience
At some time between 1.12.0 and 1.13.0, the release tools used for bindfs has changed. Before 1.13.0, the tarball name will omit the patch release number if it was 0. Since 1.13.0, the full version number is always used.
Ah, that was just me the human being inconsistent. Sorry about that :)
@mpartel No worries, that was much easier to solve than a real problem ;)
Thank you guys for this work! I updated the plugin, but unfortunately got the same error.
sudo vagrant plugin install vagrant-bindfs Installing the 'vagrant-bindfs' plugin. This can take a few minutes... Fetching: vagrant-bindfs-1.1.2.gem (100%) Installed the plugin 'vagrant-bindfs (1.1.2)'!
sudo vagrant up --provider=libvirt Bringing machine 'pricecomparer' up with 'libvirt' provider... ... ==> pricecomparer: -- Base box: centos/7 ... ==> pricecomparer: Bindfs seems to not be installed on the virtual machine, installing now The following SSH command responded with a non-zero exit status. Vagrant assumes that this means the command failed!
[ -f ./bindfs.tar.gz ] && rm ./bindfs.tar.gz
Stdout from the command:
Stderr from the command:
Hi @BoBRoID
Can you please copy your Vagranfile here?
@BoBRoID The incriminated command is nothing more than an end trap runed after the installation script to clean temporary files. I guess the real problem may hide behind.
I just released a new release, in which this command has been silenced even when it fails. Can you please update the plugin and launch vagrant up
again?
Thanks
Hi @BoBRoID
Can you please copy your Vagranfile here?
Hi @gael-ian!
I uploaded my config here: https://github.com/BoBRoID/my-vagrant-config
@BoBRoID The incriminated command is nothing more than an end trap runed after the installation script to clean temporary files. I guess the real problem may hide behind.
I just released a new release, in which this command has been silenced even when it fails. Can you please update the plugin and launch
vagrant up
again?Thanks
Yeah, I've tried this at my laptop and I've got another error. It's too large so please see it here: https://gist.github.com/BoBRoID/3d40525e9d6e7e5c6a88b0ad801819d0
@BoBRoID
Ok, my bad. I did some refactoring when I released 1.1.2 that introduce a stupid bug on source install command building. I should have unit tests for that but it's not really easy to test code that will be executed on guest machines. The only method I found is to launch tests machines and there's a lot of them...
I'll try to find a better testing method in the future but for now, your problem should be fixed with vagrant-bindfs 1.1.4. I runned a full provisioning of a CentOS 7 box with installation from source and everything went fine.
Sorry for the inconvenience
Hi all! Today I've reinstalled OS on my computer, and found that all my project which are using bindfs plugin won't able to work. I saw an error in my console:
The following SSH command responded with a non-zero exit status. Vagrant assumes that this means the command failed!
After hours of digging I realized that it might depends on plugin's version. I've downgraded it (from 1.1.1 to 1.1) and now all working correct. I guess it would be helpful, because I spend few hours to solve this error message.