Workflows and credentials are already separated in different tables but the same database. Having them in separate databases would not be a good idea.
Yes, an option which deactivates all workflows sounds like a good idea.
That it happens again was sadly to expect. If it failed before and caused problems, and you do exactly the same again, the chance that it breaks again is very very high.
That is why I mentioned that you should not activate the workflow again. There seems to be a bug. Sadly am I not able to reproduce it.
If something goes wrong I can not activate the workflow at all.
Even if I activate it manually via the database, trying to reproduce this problem, n8n simply displays an error on startup, does not activate the workflow but starts normally.
So is sadly currently impossible for me to fix this issue. Sorry.
before I did this I wasn’t sure which object is the culprit. Is there anything I could log for you to better understand what is going on?
I assume it is the google label I use in the configuration [ifttt] but I haven’t tested further.
I would be willing to do/try whatever is helpful.
I tried with default INBOX … but same thing:
There was a problem and the workflow could not be activated:
API-Server can not be reached. It is probably down.
API-Server … is not really needed with imap … don’t know what this is complaining about.
Maybe it’s a tunnel thing.
Connection to the db seems to break … is this the API-Server?
The API is the n8n backend. It is literally needed for everything as that is doing all the work.
No matter what you select as Mailbox, the problem will not go away. It has to do with the SSL certificate. The only thing which can help here is adding the option “Ignore SSL Issues” on the node and setting it to true.
Added an option to the CLI to deactivate workflows. It will be released with the next version.
In the screenshot of the node you posted, I can already see all the options. The only other thing that would be interesting is a screenshot of the credentials. With user and password blacked out. Pretty sure there is nothing surprising there but who knows.
Thanks … Ignore SSL Issues works. I ran the workflow successfully. Is there a downside I should know about?
The imap trigger is not active when I activate the workflow - I rather have to execute the workflow … then I see the imap animation and it notices the imail in [ifttt]. Good … but why isn’t “active” enough to process the workflow … do I have to press the play butten for the workflow every time I reboot the VM with n8n?
When I run the workflow again … the same thing happens. So if I want to poll the folder for a new email … I have to make sure the mail is deleted after being processed. Alternatively setting the trigger to “unseen” should also do the trick. Right?
Actually I wonder what the Trigger does if I don’t set it to unseen and use my inbox?
Attached are my credientials … which are in fact ok … since it did something now
Thanks for your patience! And good to hear about that new CLI option
Yes, the downside of ignoring SSL issues is that it is now easier for hackers to pretend they are Google. After all, it now also connects if there are issues with the certificate. Like if it is expired, can not be confirmed, is invalid, …
If you activate a workflow it runs in production-mode. So it runs the workflow whenever an email gets received in the background. You will not see anything in the editor-UI (browser). But it is active as long as n8n is running. Also if you restart the VM it will also be directly get activated again (you could see that already because of your earlier problems).
You will only see the data in the editor-UI in testing mode. So if you press “Execute Workflow” in the editor.
About your process question. That is already explained in the description field for “Action”: What to do after the email has been received. If "nothing" gets selected it will be processed multiple times.
