Nightly refresh did not complete, dashboards showing stale data
Ticketclosed
Brimscombe Energy · Nightly refresh did not complete, dashboards showing stale data
Siân Pritchard reported nightly refresh failed, leaving dashboards 2 days stale. Kate Ormerod identified Snowflake compute timeout; the platform team optimized partitioning and refresh completed successfully.
What happened
What this says
As at Fri 9 Oct 2026Siân Pritchard reported nightly refresh failed, leaving dashboards 2 days stale. Kate Ormerod identified Snowflake compute timeout; the platform team optimized partitioning and refresh completed successfully.
Problem: Nightly refresh did not complete, leaving dashboards two days stale. This was urgent because customer teams needed current data to answer questions during morning standups.
Handling: Kate Ormerod responded in 52 minutes. The platform team identified a Snowflake query timeout during nightly load due to compute spike and optimized query partitioning to prevent recurrence.
Watch: Siân confirmed dashboards show current data. This was a data freshness SLA event resolved quickly, but compute spike risk may recur if partitioning optimisation is insufficient.
Details
Updated 9 Oct 26Trail
created · Siân Pritchard 21 Jul 2026 13:16Our nightly refresh didn't finish last night. The dashboards are showing data from two days ago. This is urgent because our customer teams are trying to get answers from the morning standup and the brain isn't reflecting the latest information. They're depending on current data to answer customer questions about accounts and contracts. Can you check what happened with the refresh job?- Public reply · Kate Ormerod 21 Jul 2026 14:08Hi Siân, thanks for flagging this. I can see the nightly refresh started but stopped shortly after. I'm working with the platform team to investigate the root cause now. I'll update you as soon as I know more.
- Public reply · Siân Pritchard 22 Jul 2026 12:37Thank you. I can see the refresh has now completed and the dashboards are showing current data as of this morning. That was concerning but I'm glad it's sorted. Diolch, Siân.
- Internal note · Kate Ormerod 22 Jul 2026 13:42Root cause: Snowflake query timeout during nightly load on customer's account due to spike in compute usage. Platform team optimised query partitioning. Refresh completed successfully by 07:00 this morning. Confirmed with Siân that data is current. Close.
- Status change · 26 Jul 2026 23:59