[destination-hubspot] Batch upsert 409 Conflict crashes entire sync

[jamesg31]

Connector Name

destination-hubspot

Connector Version

0.0.10

What step the error happened?

During the sync

Relevant information

The HubSpot destination connector throws IllegalStateException when the batch upsert API returns 409 Conflict, failing the entire sync.

java.lang.IllegalStateException: Invalid response with status code 409 while starting ingestion:
{"status":"error","message":"Contact already exists. Existing ID: 100887063864","correlationId":"...","category":"CONFLICT"}

Analysis

The 409 occurs when HubSpot's batch upsert cannot find a contact by the idProperty lookup, falls back to creating a new contact, and the creation fails due to a unique property constraint.

With idProperty: email and no duplicate emails in the source data, I am unsure what is causing Hubspot to return a 409. I suspect this is just a transient issue with the Hubspot API. Either way, these failures should not fail the sync. Some possible issues could be:

• Eventual consistency under load — with ~2,460 batches, HubSpot's email lookup index may temporarily lag behind recent writes, causing the lookup to miss an existing contact while the uniqueness check on creation catches it
• hs_additional_emails collision — if a source record's email doesn't match any contact's primary email but matches another contact's hs_additional_emails, the lookup could fail while creation conflicts. I confirmed this reproduces the exact 409:

curl -X POST '.../batch/upsert' -d '{"inputs": [{
  "idProperty": "email",
  "id": "new@example.com",
  "properties": {"email": "new@example.com", "hs_additional_emails": "existing@example.com"}
}]}'
→ 409: "Contact already exists. Existing ID: 100887063864"

Expected Behavior

A single problematic record should not crash the entire sync. Failed records should be routed to the DLQ while the rest of the batch succeeds.

Relevant log output

Contribute

• Yes, I want to contribute

---

Internal Tracking: https://github.com/airbytehq/oncall/issues/11580

[octavia-bot]

AI Triage Starting...

Thank you for opening this issue! I'm an AI assistant that will do an initial review and may:

• Ask clarifying questions if needed
• Suggest relevant documentation or workarounds
• Flag this for internal visibility

Note: Community issues are reviewed on a best-effort basis and do not have a guaranteed response time. For additional help:

• Join the Airbyte Community Slack to connect with other users and community members
• If you're an Airbyte Cloud customer, you can open a support ticket or contact your account representative, referencing this issue URL

View playbook

• 🔗 Session URL: https://app.devin.ai/sessions/b336e00fb575468587f553a7689d7906

✅ Devin AI session created successfully!

[devin-ai-integration]

Hi James Gardner (@jamesg31), thank you for reporting this issue — and for the thorough analysis!

This looks like a legitimate bug in how the destination-hubspot connector handles 409 Conflict responses from the HubSpot batch upsert API. As you noted, the connector should not crash the entire sync when a single record conflicts — failed records should ideally be routed to a dead-letter queue (DLQ) while the rest of the batch proceeds.

Your investigation into the hs_additional_emails collision and the eventual consistency hypothesis under high load are both plausible root causes for the 409. The key issue is that the connector's error handling treats 409 as a fatal IllegalStateException rather than a recoverable per-record failure.

I've escalated this to the internal team for further investigation: https://github.com/airbytehq/oncall/issues/11580

---

Need more help? Join Airbyte Community Slack for peer support, or if you're a Cloud customer, open a support ticket referencing this URL.

This message is generated from Devin.ai (internal session link).