We've been running our transactional and marketing emails through Sendinblue for about 18 months, and for the last quarter, we've been plagued by a critical issue: our automations (specifically, workflow sequences) will randomly enter a "Paused" state without any manual intervention or clear system error. The automations are not hitting contact limits, and our sending reputation is consistently green. This happens across different workflow types—welcome series, abandoned cart, post-purchase nurture—with no discernible pattern.
When we contact support, the response is cyclical and unhelpful. They typically:
* Acknowledge the issue after 48+ hours.
* Manually resume the workflows on their end.
* State they are "investigating" but provide no root cause.
* Close the ticket after a week without follow-up.
We've ruled out the obvious:
1. **API Calls:** We aren't programmatically pausing via their API (`PUT /api/v3/processes/{id}/pause`).
2. **Contact Modification:** The pauses don't correlate with bulk imports or segment updates.
3. **Performance Limits:** We're far below our plan's contact and email hourly limits.
Our current workaround is a cron job that polls the Sendinblue API to check the status of key workflows and alerts us if they're paused. This is unsustainable.
```python
# Example of our monitoring script snippet
import requests
import time
WORKFLOW_IDS = [12345, 67890]
API_KEY = "your-key-here"
def check_workflow_status(workflow_id):
url = f"https://api.sendinblue.com/v3/processes/{workflow_id}"
headers = {"api-key": API_KEY}
response = requests.get(url, headers=headers)
if response.status_code == 200:
return response.json().get('status') # Returns 'paused', 'inProgress', etc.
return None
for wf_id in WORKFLOW_IDS:
status = check_workflow_status(wf_id)
if status == 'paused':
# Send alert to Slack/email
print(f"Alert: Workflow {wf_id} is paused.")
```
Has anyone else encountered this specific "random pause" behavior? More importantly, has anyone successfully diagnosed the cause or found a reliable fix beyond constant monitoring? We're considering building a middleware layer to auto-resume via their API, but that feels like papering over a fundamental platform instability.
Given the subforum theme, I'm also keen to hear comparative experiences with automation stability on platforms like Customer.io, ActiveCampaign, or even building on a lower-level service like Postmark + a workflow engine. The sync reliability with our CRM (HubSpot) is the next concern if the core automation engine is unreliable.
API first.
IntegrationWizard