Why does my email-triggered workflow keep failing with the same error message?

Help
1

This topic was automatically generated from Slack. You can find the original thread here.

Hey guys, i’ve got something strange that’s come up.
I’ve got a workflow that is triggered by email. It’s been working just fine. But for the last couple days, it fails with the same message every time

Error: Process ended too soon
    at handleError (/var/task/common.js:38:40)
    at Runtime.handler (file:///var/task/lambda_handler.mjs:989:5)
    at process.processTicksAndRejections (node:internal/process/task_queues:95:5)

Each failure shows 0/18 steps executed and the execution time is between 30 and 130 seconds. When I check the flow, nothing happens beyond the trigger step.

This really feels like an internal error, and I’m fairly sure I havent changed anything before this started breaking. Does anyone have any insight into this issue?

2

Hi , do you have any code step in your workflow. Code step is usually the cause for these kind of error

3

Yeah, the workflow is primarily Python code steps. But whenever I had code timeout in the past, I could clearly see what step was reached and why it timed out. In this case, it’s timing out on the trigger step.

4

it’s timing out on the trigger step.
Actually it’s a display issue. There’s an error in your code step.

I would recommend you to test each of your code step to pinpoint the error. Maybe some libraries you used in your code getting updated that break your code?

5

Ok, a display issue makes sense. I just checked an API that is called through my code steps and sure enough it has been called over the last couple days. I’m not sure how to troubleshoot this off the top of my head, but I should be able to figure it out from here. Thanks Leo!

6

Yeah, those issues are really hard to debug without straightforward visibility of which steps did or didn’t run. :disappointed:

7

Yeah we’ve seen some user reported their pain-point related to this issue. I’ve created a ticket somewhat related here.

Hopefully the new v3 workflow Pipedream team is working on will have some improvement on this

8

I’ll add one more note here regarding troubleshooting.

If you recall my workflow is triggered by an email. The email contains an attachment. But since the attachments are no longer available when I try to rerun the workfow, it is nearly impossible to try to recreate what happened.

Overall I love Pipedream enough that its worth the hassle, but I figured I’d mention it in case you werent aware of this issue.

9

If you recall my workflow is triggered by an email. The email contains an attachment. But since the attachments are no longer available when I try to rerun the workfow, it is nearly impossible to try to recreate what happened.
Hmm I think you can try to download a file to your workflow using the Helper action - Download to /tmp. Then update your action to use that file to test

10

Yes thats the first step after the trigger. But since only the trigger is firing, that means i need to manually run the download_to_tmp step… which doesnt work because the attachment is no longer available. Its a real pickle.

11

it’s possible that your workflow is running out of memory. Could you try increasing the memory of the workflow in its settings?