This topic was automatically generated from Slack. You can find the original thread here.
I think there’s something bugging with the hubspot-new-engagement trigger. I can deploy it alright, and it does send me an initial batch of old events. But I get nothing at all after that initial batch.
Judging from what I can see, the hubspot-new-engagement API calls a legacy endpoint from v1. So maybe it’s not reliant anymore?
Looks to me like that component should be using the activities endpoints from v3 instead, which have separate endpoints for each activity type - e.g. emails here:
Looking at the source logs, the second one you linked had exceeded memory usage on one run, so it probably fetched too many objects (and hubspot objects can be pretty large). The original one seemed to be running and just not retrieving anything.
Yeah, I noticed when I deployed the dc_gzuxvel trigger that the initial sample batch I got was event dating all the way back to 2024. So my guess is that at some point more recently, HubSpot stopped logging engagements in ways that would be picked up by that endpoint, and probably moved to what that new endpoint I shared picks up. Which would explain why the trigger is running but not receiving anything.
Regarding the second dc_x2uvgpA trigger, is there a way I can spot incidents like that myself, so I can resolve it. What do I do in such scenarios?
We did preliminary investigating into the new-engagement trigger, but I couldn’t repro any similar scenarios, and we still need to evaluate updating to the endpoint you suggested - I’ll bubble this up with our team today.
Did the fetchAllProperties: false solution only temporarily work for you?
About the other two HubSpot triggers that Nicki and Diego reported with issues - more details would be helpful for us to reproduce and investigate.
fetchAllProperties: false did resolve that issue for me, so yes, I suppose it was a memory issue. Ideally, I would like to somehow be notified when something like that happens, like a timeout or a crash, so I can respond to it with my infrastructure. A general infrastructure events webhook of sorts would be super useful.
I am still not seeing any results from the dc_gzuxvel (hubspot-new-engagement) trigger, this is still not working for me. What I can say is that it seems to only recognize events from around summer 2024, nothing newer. That is what I receive with the initial emit upon deploy, but nothing after that. So my best guess is simply that the endpoint utilized by this trigger is outdated and not compatible with Hubspot Engagements today.
What more would you need from me, to be able to zoom in on this? Happy to help!
The possible migration to the new endpoint(s) for new-engagement is on our team’s radar already. About error notifications, have you tried this? Handling Errors - Pipedream