* allow exposing meta information for registered metrics
New public method `ExposeMetadata` allows enabling exposition
of dummy meta-info for all exposed metrics across all Sets.
This feature is needed to improve compatibility
with 3rd-party scrapers that require meta information to be present.
This commit doesn't update exposition of default system/process
metrics to keep the list of changes small. This change should
be added in a follow-up commit.
https://github.com/VictoriaMetrics/metrics/issues/48
* cleanup
* wip
* wip
* wip
* wip
---------
Co-authored-by: Aliaksandr Valialkin <valyala@victoriametrics.com>
Log-based histograms provide lower estimation error for the same number of buckets compared to log-linear histograms.
For example, the current Histogram implementation splits each decimal range (10^n .. 10^(n+1)] into 18 buckets.
These buckets have the following bounds:
- for log-linear histogram: (1 .. 1.5], (1.5 .. 2], (2 .. 2.5], ... (9.5 .. 10]
- for log-based histogram: (1 .. 1.136], (1.136 .. 1.292], ... (8.799 ... 10]
The maximum estimated error for log-linear histogram is reached in the first bucket per each decimal range and equals to 1.5-1=0.5 or 50%.
The maximum estimated error for log-based histogram is constant across buckets and equals to 1.136-1=0.136 or 13.6%.
This means that log-based histogram improves histogram accuracy by up to 50%/13.6% = 3.6 times when using the same number of buckets.
Further reading - https://linuxczar.net/blog/2020/08/13/histogram-error/
