Pausing workflow after multiple errors

Questions
1

would it be possible to somehow pause a workflow after several errors in a row?

I have a workflow that starts every 1 minute to update the data.

And in case of error, I get a message.

the problem occurs when an error occurs, because every 1 minute a new message is resent because the workflow tries to process the same data with an error.

My intention was that when a workflow failed 2 runs in a row it would be paused.

That way I could fix the error and activate the workflow manually.

2

No, that is sadly currently not easily possible.

You can however theoretically build it yourself with an Error Workflow that uses the internal (and so undocumented) API to query the past executions and then also use the internal API to deactivate the workflow. Would however be a little bit more complicated and take so more time.

3

via http: //localhost:5678/rest/executions-current

returns executions in progress, is there any way to pull the execution history?

that way I would be able to analyze the time of the last execution of the error, and with an IF node not send the message

4

That would be without -current so: http://localhost:5678/rest/executions

5

I’m impressed with the amount of content in this community and also with everyone’s willingness to help.
I managed to find a solution to the problem.

I would like to leave my contribution

fThe original workflow only starts if the error flow is no longer running.

And the workflow error checks if it already has a running in progress, if it doesn’t it checks the last 2 dates of the run error and checks if the interval is less than 3 minutes. If it is, it checks the date of the last update that occurred in the flow.

If the update date is greater than the error date, it closes the error.

This way, it is possible to check if the error correction has already been done.

If not corrected it will wait 1h and check again.

This way the original stream will not run because it is waiting for the error stream to be terminated

I only ran a few tests, there may be an error, but I’m sharing if it helps someone else