On June 24th at 12:18 UTC our engineers identified an issue whereby one of our data collection servers wasn't properly writing data. 

What happened?

One of our data collection servers was not properly writing data to disk between June 23rd at 19:20 UTC and June 24th at 12:40 UTC. This could have caused about 4% of heatmap data and 4% of recordings to not be collected. 

Why did this issue occur?

The offending server experienced a database connection issue which caused it to stop processing incoming session data and therefore not writing the data to disk.

What will we do to prevent this from happening in the future?

We will hold an internal postmortem to establish any relevant follow-up actions.