Do in-progress workflows persist across reboots?

[Daniel_Mulroy]

Describe the issue/error/question

Our server is currently configured to reboot at a set time over the weekend.

We’re looking to implement a workflow that has a multi-day “wait” node, initiated by a webhook at any time (an externally driven ‘start’ event).

What happens to workflows that are ‘wait’-ing during a server reboot? Do they persist? Is there a difference between “wait 5 days” vs. “wait until a specific date” that would cause it to be resilient?

Sorry for asking this instead of testing it myself but I wanted an authoritative answer instead of repeatedly rebooting my production n8n server to test

Please share the workflow

Information on your n8n setup

• 0.199
• Running n8n via Docker stack

Thank you!

[MutedJam]

Hi @Daniel_Mulroy, I believe these should be fine. As long as the wait time is over 65 seconds, the execution data would be persisted to the database, meaning it should still be available after the reboot.

I’ll double check this and will get back to you

[Daniel_Mulroy]

That would be

Thanks @MutedJam

[MutedJam]

Yep, looking good. Started a workflow including a 10 minute wait, rebooted my instance, the workflow was still in waiting state with the execution data still available following the reboot:

It finished as expected once the waiting time was over, with the full execution data available:

Hope this helps!

[Daniel_Mulroy]

@MutedJam I could have sword I already replied, but it seems I didn’t.

Thank you so much for the authoritative answer!