This topic was automatically generated from Slack. You can find the original thread here.
Vic N : Hi guys, great tool. I have referred to the cold start documentation (and recommended solution), but still have been subject to cold start on my workflows. What’s the exact interval in which I should be s ending a warm request? I tried 15 as recommended in the documentation, even 10 but still seem to run into cold starts on my workflows.
Dylan Sather (Pipedream) : Hi , thanks for following the docs and I’m sorry this didn’t work for you. We don’t have control over the uptime of the execution environment - AWS handles this, and they don’t publicly document the max uptime when no requests have been received. In practice it’s ~15 but as you noted, the environment can spin down earlier.
Can you try a 5 min frequency and let me know if that works? That should be 288 invocations a day, which is well below the daily limit of 3,333 invocations. And if you need to warm more workflows from here, we’ll still only count a single invocation per run, no matter how many steps you have in your workflow.
Vic N : Yep! I’ve already upgrade to the pro plan so not worried about the invoke limits. 5 minutes seems to do the trick from what Ive seen so far. Thanks for the quick response!
Dylan Sather (Pipedream) : Great, I’ll update our docs accordingly.