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
The metrics data model specification has covered Histogram here.
The goal is to have the SDK specification supporting histogram:
What would be the default histogram buckets?
What type of buckets of we support (arbitrary buckets, equal size buckets, exponential buckets, etc.)? If we have exponential buckets, what base (e.g. base 2 or base 10) do we support?
What type of compression format / algorithm do we support (e.g. UDDSketch, etc.)?
Do we allow conversion / interpolation (e.g. convert base 2 buckets to base 10 buckets with potential data loss)?
Some of these topics are being discussed in #1776, and we try to resolve it next week (7/13/2021) - @jmacd FYI.
We've discussed this during the Metrics SIG meetings and decided to move "base 2 exponential histogram" out of the initial stable release scope (which means the metrics SDK spec stable release won't be blocked on base 2 exponential histogram, if we can catch the feature-freeze train we might be able to take it".
What are you trying to achieve?
The metrics data model specification has covered Histogram here.
The goal is to have the SDK specification supporting histogram:
Some of these topics are being discussed in #1776, and we try to resolve it next week (7/13/2021) - @jmacd FYI.
Related to #1260.
The text was updated successfully, but these errors were encountered: