You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right now we have a few conditions where we output NaN where we likely can do a bit better. Counter restarts is one edge case, and transitioning between two time series in the runningDifference calculation is another I believe.
I think this gets exacerbated if a service is rapidly scaling up/down as it'll create transitions between two time series, and worse of all this gets worse on bigger granularities rather than smaller I think.
fixes#120
Intended to handle the case where a nonsensical rate would be returned (e.g. massively negative) or the metric `_string_attributes` change mid-query, to either clamp to zero or filter out possible NaN being currently returned. This should create cleaner graphs for rate metrics, allowing the rate to interpolate over missing data instead of creating a gap in the graph
Right now we have a few conditions where we output NaN where we likely can do a bit better. Counter restarts is one edge case, and transitioning between two time series in the runningDifference calculation is another I believe.
Related code: https://github.com/hyperdxio/hyperdx/blob/main/packages/api/src/clickhouse/index.ts#L792-L805
From SyncLinear.com | HDX-134
The text was updated successfully, but these errors were encountered: