As discussed in this PR comment, process.uptime is currently implemented as a TimeGauge in UptimeMetrics.
Semantically, it should be a FunctionCounter since it represents a monotonically increasing accumulated value.
However, for some registries, the change from a gauge to a counter is a breaking change. So we will need to consider how to do this in the least disruptive way, or wait for 2.0 to do this.
As discussed in this PR comment,
process.uptimeis currently implemented as aTimeGaugeinUptimeMetrics.Semantically, it should be a
FunctionCountersince it represents a monotonically increasing accumulated value.However, for some registries, the change from a gauge to a counter is a breaking change. So we will need to consider how to do this in the least disruptive way, or wait for 2.0 to do this.