Hey folks, data_shipper_joe here. Usually I'm over in the data integration threads talking about syncing CRM events to a lakehouse, but today I need to vent a bit about an ops experience that really caught us off guard.
We rely on PingFederate for some critical customer identity flows. Last month, during a major product launch, we started seeing a spike in failed authentication requests—like, 30% failure rate. We opened a Sev 1 ticket with Ping support right away, as this was directly impacting live users.
The initial auto-response was fast, but then... silence. It took over four hours to get a human to even acknowledge it was being looked at. For a Sev 1, that's an eternity when you're watching your error graphs climb. We were providing logs, configuration snippets, and timeline details, but each exchange felt like it was on a 60-90 minute delay. I understand complex issues take time to diagnose, but the communication latency itself became a huge problem.
```json
// Example of the kind of error pattern we were seeing in logs
{
"timestamp": "2023-10-26T14:32:01Z",
"component": "PingFederate",
"error_code": "SAML2_ARTIFACT_VALIDATION_FAILED",
"sp_context": "Critical_Checkout_Flow"
}
```
In my world, if a core data pipeline is down, our vendors are typically on a live call with us within the hour for a Sev 1. We ended up having to brute-force a workaround by scaling instances and rerouting traffic before Ping support fully isolated the root cause (which, to be fair, was a gnarly edge-case bug on their end).
Has anyone else run into similar support response issues with Ping on critical tickets? I'm curious if this was a one-off bad day or a more common pattern. The product itself is generally solid for our use case, but this support experience has us re-evaluating our escalation procedures and maybe even our long-term vendor stability requirements.
ship it
ship it
That latency on a Sev 1 is rough, especially during a launch. Did your contract have any stated SLAs for response time, and did you get any kind of credit or explanation for the miss afterward? We're shopping for a new identity provider and support performance is a huge factor for us.