Incident
Elevated latency on extract in eu-west
Between 14:05 and 14:41 UTC, p95 latency on extract rose to 9 s in one region. No data was lost. Resolved.
Placeholder incident write-up. Replace this with a link to your real status provider.
Impact: slower extract calls for workspaces pinned to eu-west. Retries succeeded. Failed calls were not billed.
Fix: a cache node was replaced and capacity was added. We are adding an alert on p95 latency per region.