Skip to content
This repository has been archived by the owner on Feb 13, 2023. It is now read-only.

Setup fails using default composer + NFS: "Could not delete web/core/.nfs" #893

Closed
mdrmike opened this issue Sep 8, 2016 · 10 comments
Closed
Labels

Comments

@mdrmike
Copy link

mdrmike commented Sep 8, 2016

When running vagrant up -- build fails. I'm not sure how to debug? Is there a verbose flag/mode I could login to watch build?

I've tried installing 5-7 times with varied configurations -- including totally vanilla without any custom config.

Failure message
This is the failure message, or some similar variation.

  Failed to download drupal/core from dist: Could not delete web/core/.nfs0000000000743e0000000001:

Note: Seems related to #741 and #733 (but both are closed). I also ran into #837 on one variation of the build, but couldn't reproduce. (might have been a flag related to drush/composer)

Build-out does work on Drupal 8 & Drupal 7 if using drupal.drush.yml instead of composer:

Host Machine Setup

  • Ubuntu-Gnome 16.04.1 LTS 64-bit (Kernel: 4.4.0-36-lowlatency)
  • Vagrant 1.8.5 (was originally 1.8.1)
  • Ansible 2.1.1.0
  • VirtualBox 5.0.24_Ubuntu r108355 {also upgraded/downgraded VirtualBox 5.1.4 r110228 (Qt5.5.1)}
  • drupal-vm: 7cd44a7 (originally tried 9930333)
  • I do not have Composer, PHP, Drush, or any other web related packages installed on my desktop host.

Vagrant Plugins

  • vagrant-auto_network (1.0.2)
  • vagrant-bindfs (0.4.9)
  • vagrant-hostsupdater (1.0.2)
  • vagrant-share (1.1.5, system)
  • vagrant-vbguest (0.13.0)

Suggestions?

@geerlingguy
Copy link
Owner

@mdrmike - Hmm... you have all the right versions, AFAICT. I can't think of anything else to try, besides completely deleting the Drupal VM checkout and cloning again.

@mdrmike
Copy link
Author

mdrmike commented Sep 13, 2016

Thanks for feedback. Ya, re-cloning is actually where I started ... in fact, it had been working under 14.04 and I've had a problem since migrating to 16.04 (I did not run an upgrade)... and I just put off looking into. I'll continue to mess with it to see if I can isolate what's going on.

@jonnyeom
Copy link

I actually got this error as well.
I was updating Drupal Core (8.1.8->8.1.10)

Update failed (Could not delete web/core/.nfs000000000000002000000001:

I have never seen this error before. Very Strange.
Is there any additional info I could provide to help solve this?

@jonnyeom
Copy link

Actually, I only ran into this issue when I ran compose update drupal/core while I was in the vagrant ssh state. Once I ran the update command outside of the ssh state, drupal updated successfully!

@daceves
Copy link

daceves commented Sep 22, 2016

I've ran into this exact issue on other machines (AWS) outside of DrupalVM, when having the code base on a mounted drive. This seems potentially like not a DrupalVM issue.

@geerlingguy
Copy link
Owner

See: #741

Specifically: #741 (comment)

I killed the unzipping of drupal/core from cache, and now (after 15+ minutes) it's going through the next steps.

It looks like it's some sort of timeout issue with large number of file operations through Composer. I typically recommend (if at all possible) running any composer operations outside the VM, because performance is usually better and these types of errors don't occur as often.

@apina-globant
Copy link

If You use drupalvm, On Vagrantfile modifiy this line:

config.vm.synced_folder synced_folder['local_path'], synced_folder['destination'], options

by:

config.vm.synced_folder synced_folder['local_path'], synced_folder['destination'], :nfs => 'nfs', :mount_options => ['rw', 'vers=3', 'tcp', 'fsc' ,'actimeo=2'], :linux__nfs_options => ['rw','no_subtree_check','all_squash','async']

This allow, to make the NFS share asynchronous.

@kentr
Copy link
Contributor

kentr commented Apr 5, 2018

This appears to have solved the problem in my case.

Add a ~/.composer/config.json file with this content:

{
    "config": {
        "process-timeout":      600,
        "preferred-install":    "dist",
        "github-protocols":     ["https"]
    }
}

Source: https://stackoverflow.com/a/26216954

@benjamin-dk
Copy link

I had to up the process-timeout to 6000, then it worked (I also restarted the VM and upped the php_memory_limit to 2G).

@webprogrammierer
Copy link

I have the same problem. Setting process-timeout to 6000 does not solve the problem:

- Removing composer/installers (v1.10.0)                                   
    Update of composer/installers failed                                     
                                                                             
                                                                             
  [RuntimeException]                                                         
  Could not delete /var/www/xxx/vendor/composer/installers/.github

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
Projects
None yet
Development

No branches or pull requests

8 participants