ttyname failed: Inappropriate ioctl for device

[jasperf]

Seems the latest deploy failed on ttyname failed: Inappropriate ioctl for device:

[staging] warning Avoid reloading php-fpm [ï.at/avoid-php-fpm-reloading]
[staging]  error  in php-fpm.php on line 49:
[staging] run sudo systemctl reload php8.0-fpm
[staging] err mesg: ttyname failed: Inappropriate ioctl for device
[staging] err sudo: no tty present and no askpass program specified
[staging] exit code 1 (General error)
task deploy:failed
task deploy:unlock

Any ideas how it could fail?

Here our yaml file

import:
  - recipe/laravel.php
  - contrib/php-fpm.php
  - contrib/npm.php

config:
  application: 'site-app'
  remote_user: forge
  deploy_path: '~/{{hostname}}'
  repository: 'git@github.com:app/site-app.git'
  php_fpm_version: '8.0'
  keep_releases: '10'
  shared_files: 
    - '.env'
    - '.transip_private_key'
    - 'storage/app/exact.api.json'
  shared_dirs:
    - 'bootstrap/cache'
    - 'public/uploads'
    - 'public/published'
    - 'storage/framework/cache'
    - 'storage/framework/sessions'
    - 'storage/framework/views'
    - 'storage/logs'
    - 'storage/tls'
    - 'storage/app/public'
    - 'storage/app/modules'
  writable_dirs:
    - 'public/uploads'
    - 'public/published'
    - 'storage/framework/cache/data'
    - 'storage/logs'
    - 'storage/tls'
    - 'storage/app/public'
    - 'storage/app/modules'

hosts:
  prod:
    hostname: 'app.com'
  staging:
    hostname: 'staging.app.com'

tasks:
  deploy:
    - deploy:prepare
    - deploy:vendors
    - artisan:storage:link
    - artisan:view:cache
    - artisan:config:cache
    - artisan:optimize
    - artisan:migrate
    - artisan:queue:restart    
    - artisan:horizon:terminate
    - deploy:publish

after:
  deploy:symlink: php-fpm:reload
  deploy:failed: deploy:unlock

Did see we had

# Allow members of group sudo to execute any command
%sudo   ALL=(ALL:ALL) ALL

forge ALL = NOPASSWD: /etc/init.d/nginx
forge ALL = NOPASSWD: /etc/init.d/php7.4-fpm restart
# See sudoers(5) for more information on "#include" directives:

#includedir /etc/sudoers.d

So updated to 8.0 and reran the script but no success as of yet. sudo systemctl reload php8.0-fpm on the server did work as said.

Command failure

Command ran from this PHP FPM file in Deployer:

....
## Configuration
- `php_fpm_version` – The PHP-fpm version. For example: `8.0`.
- `php_fpm_service` – The full name of the PHP-fpm service. Defaults to `php{{php_fpm_version}}-fpm`.
- `php_fpm_command` – The command to run to reload PHP-fpm. Defaults to `sudo systemctl reload {{php_fpm_service}}`.
## Usage
Start by explicitely providing the current version of PHP-version using the `php_fpm_version`.
Alternatively, you may use any of the options above to configure how PHP-fpm should reload.
Then, add the `php-fpm:reload` task at the end of your deployments by using the `after` method like so.
php
set('php_fpm_version', '8.0');
after('deploy', 'php-fpm:reload');

 */
namespace Deployer;

set('php_fpm_version', function () {
    return run('php -r "printf(\'%d.%d\', PHP_MAJOR_VERSION, PHP_MINOR_VERSION)"');
});
set('php_fpm_service', 'php{{php_fpm_version}}-fpm');
set('php_fpm_command', 'sudo systemctl reload {{php_fpm_service}}');

desc('Reloads the php-fpm service');
task('php-fpm:reload', function () {
    warning('Avoid reloading php-fpm [ï.at/avoid-php-fpm-reloading]');
    run('{{php_fpm_command}}');
});
...

It failed at php_fpm_command so it ran sudo systemctl reload {{php_fpm_service}} or sudo systemctl reload php8.0-fpm and on server I also ran sudo systemctl reload php8.0-fpm so... what is the issue?

New Custom Command?

I had

run('sudo /etc/init.d/php8.0-fpm restart'); // Using SysV Init scripts

for PHP command before. Perhaps we need this yaml style somehow.However, reading https://askubuntu.com/questions/911525/difference-between-systemctl-init-d-and-service (systemctl vs init.d) it seems like what I have should work?!

[jasperf]

I saw script was calling local PHP 7.4

[staging] sudo: no tty present and no askpass program specified
[staging]  error  in php-fpm.php on line 49:
[staging] exit code 1 (General error)
task deploy:failed
[staging] /usr/local/Cellar/php@7.4/7.4.25/bin/php /Users/jasper/code/site.com/site.com/vendor/deployer/deployer/bin/dep worker --port 58665 --task deploy:failed --host staging --decorated -vvv

so switched to 8.1 and tried again. It failed. Then I tried with local PHP 8.0. It still failed again as well.

Using git_tty: true does not seem to help either.

and we already have

cat /root/.profile 
# ~/.profile: executed by Bourne-compatible login shells.

if [ "$BASH" ]; then
  if [ -f ~/.bashrc ]; then
    . ~/.bashrc
  fi
fi

mesg n || true

so the possible solution mentioned at https://programmerah.com/linux-error-ttyname-failed-inappropriate-ioctl-for-device-solution-2134/ does not fly either.

[jasperf]

Solved with new visudo command for passwordless sudo:

forge ALL = NOPASSWD: /bin/systemctl reload php8.0-fpm

See #2857

[antonmedv]

Avoid reloading php-fpm https://ï.at/avoid-php-fpm-reloading

[jasperf]

True, but I have not found the way to apply this on a Forge setup server. Do recall us discussing this somewhere here before. Will need to look into this again. Checking your pointers such as

+ fastcgi_param SCRIPT_FILENAME $realpath_root$fastcgi_script_name;

whereas we have

fastcgi_param SCRIPT_FILENAME $realpath_root$fastcgi_script_name;

now on staging for testing. Did have to reload PHP FPM to reload the symlink path as well it seems...?

On production this line is missing. This as we want to wait some more.

This and I did check for script filename output on staging as well just now:

php -i |grep SCRIPT_FILENAME
$_SERVER['SCRIPT_FILENAME'] =>

some more. Just need to make sure that all is well post possible updates.

[jasperf]

perhaps this

location ~ \.php$ {
        fastcgi_split_path_info ^(.+\.php)(/.+)$;
        fastcgi_pass unix:/var/run/php/php8.0-fpm.sock;
        fastcgi_param DOCUMENT_ROOT $realpath_root;
    	fastcgi_param SCRIPT_FILENAME $realpath_root$fastcgi_script_name;
        fastcgi_read_timeout 300;
        include fastcgi_params;
    }

is off somehow..

Checking https://lorisleiva.com/deploy-your-laravel-app-from-scratch/deploy-using-laravel-forge some more as well

[antonmedv]

Can’t tell how to fix it on forge.

In Deployer provision recipe, server configured correctly out of the box.