This topic was automatically generated from Slack. You can find the original thread here.
I keep on getting timeouts when trying to execute actions for Hubspot (i.e. hubspot-search-crm). Is thtat a known issue?
Timeouts or rate limits?
Could you share a screenshot maybe?
It’s via API. So I just get an ESOCKETTIMEDOUT error from my client.
Could you share your code calling the API?
Yup, here’s the payload:
{
"id": "hubspot-search-crm",
"external_user_id": "1722883882380x262440767981062980",
"configured_props": {
"hubspot": {
"authProvisionId": "apn_nphe1vg"
}, "objectType": "contact",
"searchProperty": "email",
"searchValue": "[mikael@mybizz.ai](mailto:mikael@mybizz.ai)",
"filters": [{"propertyName":"hs_createdate","operator":"GTE","value":"2025-08-07T00:00:00Z"}],
"exactMatch": false
}, "dynamic_props_id": "dyp_Y8U2al3"
}
And could you share your client code as well? We would like to reproduce it.
Also is this also happen for other actions?
Wait, what client code do you need, Leo? You mean the SDK initialization? Mikael, can you
- Try the same configuration in the test app? One SDK, thousands of API integrations
- Paste the full response you’re getting?
Yes I’m asking for the the Pipedream SDK initialization code. Sorry for not being clear though.
the demo page doesn’t allow me to add the filters object to the call. So I had to submit it without that part and got a perfectly fine response back:
{
"os": [],
"ret": [
{
"id": "89400770970",
"properties": {
"address": null,
"annualrevenue": null,
"city": "Austin",
"company": "[mybizz.ai](http://mybizz.ai)",
"country": "United States",
"createdate": "2025-01-06T12:05:47.294Z",
"email": "[mikael@mybizz.ai](mailto:mikael@mybizz.ai)",
"fax": null,
"firstname": "Mikael",
"hs_createdate": null,
"hs_email_domain": "[mybizz.ai](http://mybizz.ai)",
"hs_language": null,
"hs_object_id": "89400770970",
"hs_persona": null,
"industry": null,
"jobtitle": "CEO & Founder",
"lastmodifieddate": "2025-09-05T18:44:54.674Z",
"lastname": "Lemberg",
"lifecyclestage": "",
"mobilephone": null,
"numemployees": null,
"phone": null,
"salutation": null,
"state": "Texas",
"website": "http://mybizz.ai",
"zip": null
},
"createdAt": "2025-01-06T12:05:47.294Z",
"updatedAt": "2025-09-05T18:44:54.674Z",
"archived": false
}
],
"exports": {
"$summary": "Successfully retrieved 1 object(s)."
},
"t": {
"br": 1757124611232,
"bp": 1757124611232,
"bls": 1757124611232,
"ble": 1757124615280
}
}
Hey , checking the payload in the demo’s runAction call (see below), I see that it is quite different from your payload:
{
externalUserId: "demo-266571f0-7194-4d03-8c2b-362affaa9757",
actionId: "hubspot-search-crm",
configuredProps: {
objectType: {
__lv: {
label: "Contacts",
value: "contact"
}
},
searchProperty: "hs_createdate",
searchValue: "2025-08-07T00:00:00Z",
hubspot: {
authProvisionId: "apn_DphWxg9"
}
},
dynamicPropsId: "dyp_d6ULalj"
}
Specifically, there’s no filters property. Additionally, hubspot error shows that Property hs_createdate is not a searchable property of object type contact(image below)
Next steps:
• Could you share how you get/define the filters prop?
• Could you share your usecase? Maybe we can implement another Hubspot action that fits your usecase better
You’re right - the filters object does seem out of place there. I suspect that must be something my LLM hallucinated when it built the query. But removing it from the payload unfortunately did not resolve the issue, as I’m still getting timeouts on my end.
FYI - I am not utilizing the SDK. I am connecting directly to Pipedream API via Bubble.io API connectors. I’m folloiwng the guide at Executing Actions - Pipedream for implementation.
Ok, I think I found the cause of the hang here.
It seems that if I set exactMatch: true or remove it completely, then I don’t get timeout. But submitting a query with exactMatch: false causes it to time out.
Got it . Thank you for your information.
This is interesting… I’ll try to reproduce this. Seems like an issue
Yeah, it sort of makes sense, as the default for exactMatch is false, so strictly speaking it’s not necessary to set it. But I wouldn’t expect it to cause any issues if I did. 
This is interesting… On my side, the pipedream connect demo works for exactMatch = false
The different is that the connect demo uses Pipedream Connect SDK, and you’re using Pipedream Connect API.
How about switching to Pipedream Connect SDK for now, so it will work for you right now?
In the mean time, I’ll try to reproduce this with Pipedream Connect API
I couldn’t reproduce using Pipedream Connect API either.
Here’s my sample Pipedream Connect API cURL that works. Maybe you can paste this to your LLM for it to update your existing code?
curl 'https://api.pipedream.com/v1/connect/actions/run' \
-H 'accept: **/**' \
-H 'authorization: Bearer XXX' \
-H 'content-type: application/json' \
-H 'x-pd-environment: production' \
--data-raw '{"external_user_id":"demo-ec4476e0-1d08-4989-8d80-612bbfb00e01","id":"hubspot-search-crm","configured_props":{"objectType":{"__lv":{"label":"Contacts","value":"contact"}},"hubspot":{"authProvisionId":"apn_V1h0eQb"},"exactMatch":false,"searchProperty":"firstname","searchValue":"PD TEST"},"dynamic_props_id":"dyp_GZUw576"}'
Hey - sorry for the delayed response.
I am testing now with your exact configuration and I am still getting esocket timeouts.
Removing the exactMatch prop or setting it to true resolves it. So it does still reproduce on my end.
I wonder if this might depend on the size of the CRM? Thinking maybe the Hubspot instance you’re testing against only has few records? The one I test against has thousands.
Happy to share an access token with you if you want to try and reproduce directly against my environment.
I found a case now that I was able to reproduce in the SDK.
Here I am making a simple search for deals where pipeline is default. It was stuck in pending state for several minutes and ultimately returned this error.
Hey , thank you for your information. This is interesting…
I’ll check with the team to fix this. Thank you 