pipedream

pipedream copied to clipboard

Is your feature request related to a problem? Please describe.

In the case my deployed workflow fails multiple times, after I fix the issue I would like to easily replay all of these events.

Describe the solution you'd like

A similar button to the DELETE ALL EVENTS as well as a way to easily bulk select all events with a failed status to replay.

Do you have a workaround? Clicking each failed event individually, then clicking Resend Event individually is the only option currently.

+1 for this. As part of this feature, it would be great to see:

• option to replay only failed steps in each event
• option to replay multiple times with exponential backoff

This would make it a much easier sell for us to move off of Zapier, which already has these abilities. I prefer Pipedream, but without this feature I'm afraid we're stuck on Zapier because of dependencies on APIs that occasionally throttle us and trigger failed automations.

@dobsonjon once bulk retry is implemented I think you could solve for those use cases with the existing $.flow.rerun feature, by editing the failing workflow to add the required $.flow.rerun logic, then using this proposed bulk replay to manually restart.

---

Can I request, in addition to:

easily bulk select all events with a failed status to replay

that we can also select all events within a time range?

I have a case where a workflow was reporting success ✅, but a bug in the code meant it wasn't doing what it was supposed to be doing. I can also imagine cases like this would benefit from a manual bulk retry because even if I had implemented $.flow.rerun logic in my workflow, it's possible a bug could result in it not rerunning even when it should.

+1 for this. Happens to me all the time.

Cruel, I just discovered the Event History page.

Now I have to decide whether to try and write an automation, or click 100s of times 😭

Just needed this again, but for several hundred failed jobs. Played around a bit with Postman trying to automate via the API, but gave up trying to script pagination.

Instead, here's my budget workaround:

Install: https://getautoclicker.com/docs/3.x/getting-started/download/

Configure like this:

And:

Then you can just paginate through clicking > and hitting Alt+A to resubmit jobs.

Would still welcome it as a feature, but at least this way can replay few hundred events in a minute or so.

+1 for being able to bulk select and retry multiple events in one go.

I would also like the Event History page to reflect which events have been retried! Currently it's very easy to forget which ones have been retried and to accidentally do the same one multiple times. My best method for avoiding this at the moment is to delete each event immediately after retrying it. But this is not ideal because I have to sacrifice my history. And, every time I delete an event, the history screen resets to Page 1 so I have to click multiple times on the Next Page button to get back to where I was working.

Oh yeah, also, it'd be great if there was an option to skip automatic retries when manually retrying an Event.

At the moment, if I retry an Event on a Workflow that has auto-retries enabled and the rerun hits another error, then it keeps retrying and erroring over and over and there's no way to stop it AFAIK. Even disabling auto-retries on the Workflow at this point doesn't stop the process. I think the only way to prevent it at the moment is to disable auto-retries on the Workflow in advance of manually retrying an Event.

+1 to "reflect which events have been retried!"

Or even replaying the workflow from the failed step, and updating the status afterwards. <-- This is how Zapier works.

So a failed event can become successful if manually replayed (from the failed step) without error.

A suggestion from an user while we wait for this feature:

Log or save externally steps.trigger.context.trace_id to later check the status / execution in the Event History, e.g. https://pipedream.com/@/event-history/<trace_id>

++1 from me. event history replay option for selected failed workflows, default option to replay from failed step only. important especially, as the "auto retry on errors" currently does not work for timed out steps, but this is the most frequent error I have when partner systems are down.