0

My shell script is not running using cron tab. If I execute this script using sudo ./myscript.sh, everything works.

myscript.sh :

#!/bin/bash

if ! cd /var/www/myfolderapps/myfolderapp; then
  echo 'Failed to cd to /some/dir' >&2
  exit 1
fi

mv cms/configs/database.json ./
mv cms/configs/options.json ./
mv cms/configs/analytics.json ./
mv cms/sql/patchs.json ./
mv cms/public/uploads ./
mv cms/node_modules ./

rm -rf cms
git clone [email protected]:****/myproject.git
mv myproject cms

rm -rf ./cms/configs/database.json
rm -rf ./cms/configs/options.json
rm -rf ./cms/configs/analytics.json
rm -rf ./cms/sql/patchs.json
rm -rf ./cms/public/uploads

mv ./options.json ./cms/configs
mv ./database.json ./cms/configs
mv ./analytics.json ./cms/configs
mv ./patchs.json ./cms/sql
mv ./uploads ./cms/public
mv ./node_modules ./cms

cd cms

node sql/index.js

node scripts/mergeOptions.js

rm -rf ./tasks
rm -rf .bowerrc
rm -rf .eslintrc.json
rm -rf .gitignore
rm -rf .scss-lint.yml
rm -rf bower.json
rm -rf changelog-template.hbs
rm -rf CHANGELOG.md
rm -rf gulpfile.js
rm -rf README.md

cd ../

/usr/local/bin/pm2 restart process.json

exit;

My app is running into another folder :

Cloning into 'myproject'...
DEPRECATION WARNING: The logging-option should be either a function or false. Default: console.log
[PM2] Applying action restartProcessId on app [myfolderapp](ids: 0)
[PM2] [myfolderapp](0) ✓
┌─────────────────────────┬────┬──────┬──────┬────────┬─────────┬────────┬─────┬───────────┬──────┬──────────┐
│ App name                │ id │ mode │ pid  │ status │ restart │ uptime │ cpu │ mem       │ user │ watching │
├─────────────────────────┼────┼──────┼──────┼────────┼─────────┼────────┼─────┼───────────┼──────┼──────────┤
│ myfolderapp │ 0  │ fork │ 5236 │ online │ 137478  │ 0s     │ 0%  │ 15.3 MB   │ root │ disabled │
└─────────────────────────┴────┴──────┴──────┴────────┴─────────┴────────┴─────┴───────────┴──────┴──────────┘
Use `pm2 show <id|name>` to get more details about an app
Improve this question
9
  • 1
    Any error messages sent to the specified email address or logged elsewhere? Is the script executable? Does it try to use environment variables that are not available? Could you say something about what the script is doing and what you expect it to do? Commented Nov 5, 2020 at 16:42
  • I dont receive any emails and if I execute the script with the command sudo ./rupdate in my folder /var/www/website everything works. Commented Nov 5, 2020 at 16:43
  • 2
    Why are you setting HOME to /? First why are you setting HOME at all, is it needed? And second, if it is needed, shouldn't that be root's home directory, /root? Commented Nov 5, 2020 at 16:44
  • 1
    Is is really webstite ? Commented Nov 5, 2020 at 16:53
  • 1
    @tonymx227 we can't know this since you don't show your script so we have no idea what it is doing. What I can tell you is that it doesn't make sense to set HOME to /. Commented Nov 5, 2020 at 17:04

1 Answer 1

Reset to default
1

Your script appears to assume that it is executed with a particular directory as its current directory. Make sure that the current directory is set appropriately at the start of the script by using cd.

For example, if the cms directory is located in /some/dir, then use something like

#!/bin/bash

if ! cd /some/dir; then
    echo 'Failed to cd to /some/dir' >&2
    exit 1
fi

# rest of script goes here

This would terminate the script with a diagnostic message if the /some/dir directory does not exist or if the current user can't cd into it for whatever reason.

If your git command (or any other utility) needs to access files in the user's home directory (for e.g. authentication), then setting HOME to a directory which is not the user's home directory would make that command fail or fail to work in the expected way. The HOME variable is typically already set correctly when executing cron jobs, and seldom have to be modified.

If you are required to modify the PATH variable to find the git, node, or pm2 commands (or any other non-standard command), then I would do so in the script itself rather than in the crontab. Setting PATH in the crontab would modify the PATH for all cronjobs scheduled by that tab, which in many cases may not be wanted.

Judging from the last line of your error output, at least the pm2 command can't be found in the default PATH, so adding the directory where pm2 is located to PATH would be necessary.

Your script also appears to expect that the cms directory exists, and it obviously tries to move some configuration files from it to a new checkout of the directory from some Git repository. It's unclear what you expect to happen if the cms directory is not found, or why you don't just do an git up in the cms directory.

The Git repository that you check out is called Gestiaweb, but the script expects a directory called Website. It's unclear whether these names are related or if Website is a pre-existing directory.

Improve this answer
3
  • Ok so now, my cron tab works but the script seems not working as expected. If I run the script manually it works but with crontab, I get an error. See edit. My cms seems to be deployed on another directory and not in the /var/www/cms/WEBSITEFOLDER directory. Commented Nov 6, 2020 at 9:49
  • Hello, I updated my post to be more clearly. Commented Nov 10, 2020 at 16:42
  • @tonymx227 It appears as if you have modified your question in such a way that my answer now is irrelevant. Instead of doing that, what I would suggest that you do is to roll back the question to the state where it contains the error log, and then ask a completely new followup question. I think this would be best as your follow up question is totally separate from your initial question. Commented Nov 10, 2020 at 18:42

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.