Allow histograms to specify negative bucket bounds #156
Comments
|
Should this be covered by the spec? |
|
Please draft a PR. |
I think of this as an OTLP issue, not a spec issue. |
x13n
added a commit
to x13n/opentelemetry-proto
that referenced
this issue
Jul 3, 2020
x13n
added a commit
to x13n/opentelemetry-proto
that referenced
this issue
Jul 3, 2020
x13n
added a commit
to x13n/opentelemetry-proto
that referenced
this issue
Jul 6, 2020
|
I drafted a PR, please take a look: #163 |
bogdandrutu
added a commit
that referenced
this issue
Jul 17, 2020
Fixes #156 @nilebox @bogdandrutu @jmacd Co-authored-by: Bogdan Drutu <bogdandrutu@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
According to the proto documentation, histogram bucket bounds can only be positive, which makes it impossible for OpenTelemetry to be fully compatible with Prometheus, which allows negative bounds as well.
Prometheus receiver doesn't respect this and sends negative bucket bounds down OpenTelemetry collector pipelines anyway. This can lead to bugs, as processors/exporters can make incorrect assumptions about the data (in fact, at least Stackdriver exporter already does).
This is a request to extend OpenTelemetry data model to treat first histogram bucket as
(-inf, buckets[0])instead of[0, buckets[0])./cc @bogdandrutu @rghetia @nilebox @mayurkale22
The text was updated successfully, but these errors were encountered: