Bugzilla

Chris Peterson discovered starting July 19th, that [`wr.scenebuild_time`](https://dictionary.telemetry.mozilla.org/apps/fenix/metrics/wr_scenebuild_time) was showing a sharp increase in Invalid Overflow errors.

Also starting at the same time was a [sharp increase in `null` values for `client_info.device_model`](https://dictionary.telemetry.mozilla.org/apps/fenix/metrics/wr_scenebuild_time).

Not certain if these are correlated by anything other than the date that this occurs as of yet. According to Chris, there were no Fenix releases on or around the time of the change in telemetry, and this was only seen in release.

One thing I did check was if there were any Nimbus experiments launched at this time and there was no experiment launch near the timeframe of the change.

Chris Peterson discovered starting July 19th, that [`wr.scenebuild_time`](https://dictionary.telemetry.mozilla.org/apps/fenix/metrics/wr_scenebuild_time) was showing a sharp increase in Invalid Overflow errors.

Also starting at the same time was a [sharp increase in `null` values for `client_info.device_model`](https://sql.telemetry.mozilla.org/queries/87013/source#215604).

Not certain if these are correlated by anything other than the date that this occurs as of yet. According to Chris, there were no Fenix releases on or around the time of the change in telemetry, and this was only seen in release.

One thing I did check was if there were any Nimbus experiments launched at this time and there was no experiment launch near the timeframe of the change.

Starting on July 19th there was an [increase in `null` values for `client_info.device_model`](https://sql.telemetry.mozilla.org/queries/87013/source#215604).

No Nimbus experiments launched at this time.