[Improvements]: Scheduled tasks

[andrasbacsai]

Scheduled tasks added here, but there are a few improvements that will be added:

• Add email notifications for success/fail
• Trim executions to last ~20 or so to limit db growth for historic runs
• Execution time limits
• Display next run time, total execution time
• Probably some standardisation/consistency issues in here

[gausie]

I'd also love to be able to control the Scheduled Tasks as IAC by having a .crontab or similar file in the project's codebase. Is this a useful feature to track here or should it be considered separately?

[peaklabs-dev]

• Also consider adding Slack notifications for failed tasks.
• Add tasks globally to coolify so for the coolify instance there can be cron jobs and for each container individually
• The pop-up input is too small for long cron task, consider using a wider area on the dashboard.
• Add a Timzone setting so we can select the timezone for each cron -> so it runs when we expect it

[gausie]

A task of mine seems to have got stuck a few days ago and there's no way to cancel it, nor did the task re-attempt the next day. Could we supply a timeout or something?

[aspantel]

We don't see the output of our scheduled tasks in our log drains. Is this not available?
Any ETA for this?

We use scheduled tasks as health checks and getting their output is essential!

[jonnyscholes]

My "Recent executions" seem to change order on page load. This sometimes means the most recently executed task (even if it's currently running) don't show up in the list. Does this sound right?

Would you be open to a PR that ensures the recent executions are the 20 most recent ones?

[martmalo]

Also want to mention that failed cron jobs shows logs once I click on them, but successful ones does not. is this meant to be like that?

Edit:
Did not notice that someone mentioned this already.

[Blochergroup]

Is there any plan to integrate this with the log drains?

[HashimTheArab]

Is it possible to cancel stuck tasks?

[AlejandroAkbal]

Yes please, and set a maximum run time

[aspantel]

@andrasbacsai where does the container's FluentBit collect logs from? I am thinking maybe we can have a custom logback config to send our scheduled task logs to the container's FluentBit somehow. Can you share the default FluebtBit config when drians are configured? Thanks

[aspantel]

@andrasbacsai we have long running tasks and Coolify does not let them finish.
Is there some healthcheck mechanism for these tasks that our tasks are getting terminated by Coolify?

[da411d]

Is there way to set timezone for cron?

[pencilcheck]

Running long running tasks sometimes will not show output at all, and the status wouldn't reflect the process being called.

[pencilcheck]

I also noticed that scheduled tasks will run multiple instances of the same command at the same time in parallel which is causing a lot of problem as some command I want to only run once.

Is there a way to disable parallel running of the command? it is very annoying to find out the hard way as it slow the VPS to a crawl

[Rustypredator]

Also Essential for some applications would be to specify the user that executes the cron inside the container.
For example Many PHP Applications such as Nextcloud etc. will need their cronjobs to be run by "www-data".
I think this should be fairly easy by adding "-u {user}" on the docker exec command.

[Rustypredator]

Its not really as integrated as i would wish but this works (at least for nextcloud):

• su www-data -s /bin/sh -c "php cron.php"

It is necessary here to supply -s /bin/sh since www-data by default does not have a login-shell. the -s parameter tells the session which shell to use.
Hope this helps someone 😉

[da411d]

Few weeks ago I had bug, when log file was created from cron context instead of http-server context, and then it was not writable for http-server context.