I’m encountering a persistent problem with a flow scheduled to run every 1 minute. Despite configuring the recurrence trigger correctly, the flow does not execute as expected. Here are the details:
Trigger Configuration: The recurrence trigger is set to run at one-minute intervals. I’ve verified that the interval, frequency, and time zone settings are correct.
Flow Execution Duration: It appears that if a flow execution takes longer than 1 minute, subsequent triggers may either overlap with the current run or not fire at all, leading to inconsistent behavior.
I have attached an image of the current trigger configuration for additional context. Has anyone experienced a similar issue or can offer further troubleshooting advice?
Add option the schedule to trigger the missing times (instead of skipping them).
Guarantee that jobs run within 1 minute.
The second point is technically challenging because we can’t precisely predict traffic spikes. We keep adding more servers (we added more today), but what happens is that tasks are scheduled and wait until a worker is available to pick them up and sometime there is no free ones due to spike.
We can probably bring the delay down to 2–3 minutes, but guaranteeing less than that is difficult.
Until about a week ago, everything was working perfectly. If we set a scheduled trigger to run every 1 minute, it would execute on time. If we added a delay of 5 minutes (or any other value), it would run accordingly. Sometimes there was a slight delay of about a minute, which was acceptable.
But right now, the situation is unacceptable.
I have flows with clients where a 1-hour delay must be respected, and they’re being processed 18 hours later. And I have several similar cases.
With triggers, it’s even more critical. If we set a trigger to run every minute, it must run at that interval. Otherwise, what’s the point of setting it at all? It used to work fine — please fix whatever was changed. Triggers are a core part of the system, and if they fail, everything fails.
The same goes for delays: if we configure a delay, the defined timing must be respected. Extending it 15–18x is not acceptable.
Let me confirm my understand, the issue with tremendous delay is now fixed, It should be back to run within 2-3 minutes (In rare case where the server crash it might takes up to 10 minutes to recover but no more than that), and from what I see from the screenshot you have posted It seems running within 2-3 minute.
Do you still face any issue where it get processed hours later after above comment?
Scheduled triggers with specific intervals must be handled with precision. In my case, any failure or delay in trigger execution is unacceptable. If I configure it to run every 1 minute (or X minutes), it must execute exactly at that interval — no more, no less.If the minimum reliable interval is now 3 minutes, then you should disable the 1-minute and 2-minute options altogether. Right now, I have several flows failing because they are not triggering at the expected times — and that’s unacceptable.
Delays inside flows are not being respected. When I set a delay of, for example, 3 hours, the execution happens after 9 hours or more — it seems like the delay is being multiplied. This is a serious issue.You clearly have a problem with timing in the system, and it’s affecting core functionality. Please provide a solution as soon as possible — this is a fundamental part of the platform and must be fixed immediately.
