Posting here because their support team is apparently on a sabbatical. Anyone else hitting consistent 5xx errors when using their batch API for CRM sync jobs?
Our setup:
- Standard batch endpoint, under documented rate limits.
- Payload is JSON, well-formed, validated.
- Errors are completely inconsistent. 502, 503, 504. Sometimes fails on job submission, sometimes mid-process.
What we've tried:
- Exponential backoff with jitter. No change.
- Reduced batch size from 1000 to 100 records. Same errors.
- Verified network path. It's not us.
The kicker? The failed jobs don't always show up in the dashboard. No trace. Support ticket opened 3 weeks ago, got the "engineering is investigating" auto-reply, then radio silence.
Patterns we suspect:
- Their internal queue system is overloaded or misconfigured.
- Silent failures when hitting certain data combinations (maybe field length?).
- They're probably throttling heavy users but not admitting it.
Need to know if this is systemic or just us. If you've found a workaround or gotten a real response, share. Bonus points if you have a contact that actually responds.
If they offer a 'free demo', you're the product.