What happened
What this says
Current to 12 Oct 26Nightly refresh failed to complete, leaving dashboards with two-day-old data. Platform incident escalated to Head of Platform; resolved by 23 Jul 2026.
Problem: Dashboards displayed data from two days prior after the nightly refresh failed to complete. Connectors appeared normal and no system changes had been made.
Handling: Olivia Grant reported a platform-wide incident affecting the refresh job and escalated to Ewan Pryce, Head of Platform. Investigation found a schema change in the customer's staging table wasn't handled gracefully by the refresh logic; the job was rolled back and re-run.
Watch: Incident resolved by 23 Jul 2026 under inc_2026_07_refresh.
Details
Updated 9 Oct 26Trail
- created · Ryan Heslop 21 Jul 2026 09:57Morning, our dashboards are showing data from two days ago. The nightly refresh didn't complete properly last night, or at least the latest data isn't in the brain. Everything looks normal on our end: connectors are green and we haven't made any changes. This is affecting everyone right now. Cheers, Ryan.
- public_reply · Olivia Grant 21 Jul 2026 12:42Hi Ryan, thanks for reporting. We've had a platform incident affecting the nightly refresh job. Our team is working on it now. I'll update you as soon as I have more information. Olivia.
- public_reply · Ryan Heslop 22 Jul 2026 12:35Understood. Do you have an estimate for when the refresh will be back up and running? Our team is waiting on the latest data to review this afternoon.
- internal_note · Olivia Grant 22 Jul 2026 16:39Escalated to Ewan. Refresh job hung during Snowflake ingestion. Root cause: a schema change in a customer's staging table wasn't handled gracefully by the refresh logic. Being rolled back and re-run. Should be resolved within the hour.
- status_change · 27 Jul 2026 23:59