Why is My Nightly Cron Workflow Consistently Failing Due to Cold Starts Eating into the Execution Budget?

Help
1

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

Hey ! I opened a support ticket about 2 weeks ago from reece.washington@calciumco.com and wanted to follow up here. I’m on the calciumco workspace (Connect plan) and have a nightly cron workflow that’s consistently failing due to cold starts eating into the execution budget. It either times out before a single step runs (0ms, 0/5 steps) or makes it into the sync step but runs out of the 750 second limit because the cold start consumed too much time upfront. Manual and replayed runs always work fine which confirms it’s not a code issue. Warm workers would solve it but that’s locked behind Business. Any help would be appreciated!

2

How much time is the cold start ‘eating up’? That doesn’t sound right

3

I see your support ticket - it was not lost, we’re just facing more requests than usual.

I assume the only trigger in the workflow is the CRON?

When you say it makes it into the sync step, do you mean it completes one step or just starts it?

4

Yes the only trigger is the CRON. When I say it makes it into the sync step I mean it starts executing but doesn’t complete, it gets partway through processing employees and then hits the 750 second limit. As for how much time the cold start is eating up, I honestly don’t have an exact number since the failed runs don’t show step level timing. I just know it consistently times out on scheduled runs but never on manual runs or replays, so the cold start was my best guess as to why. Happy to share any execution IDs or other details that might help narrow it down.