On 2020-10-19 at 15:20 UTC our engineers identified an issue with data collection that affected Recordings, Heatmaps and Forms.
From 14:40 UTC until 16:00 UTC we failed to collect Recording, Heatmap and Forms data for a quarter of all visitor sessions. It was random which visitor sessions failed to record data.
Why did this issue occur?
The issue was the result of us launching more data ingestion servers in response to higher traffic. The new servers were misconfigured, running an outdated version of key services.
What will we do to prevent this from happening in the future?
We will be establishing a better process around launching ingestion servers and ensuring they are working correctly.