Clarify in Composite Samplers OTEP the unreliable threshold case#4569
Conversation
Addressing issue open-telemetry#4566.
|
We usually don't update OTEPs, and instead clarify or refine things in the actual Spec contents. Probably worth checking this clarification exists in the Spec (too)? Wdyt @jmacd ? |
|
I would feel comfortable adjusting this in the specification directly. It seems like a very small change to write a new OTEP. |
|
This PR was marked stale due to lack of activity. It will be closed in 7 days. |
|
Closed as inactive. Feel free to reopen if this PR is still being worked on. |
|
@kentquirk @jmacd @oertl ping ;) |
jmacd
left a comment
jmacd
left a comment
There was a problem hiding this comment.
This is excellent, thank you @PeterF778. I will recount that over 4 OTEPs on the topic of sampling and as many years, we've repeatedly searched for how to represent something I have ill-described as "zero adjusted count". The idea that we let spans through a probabilistic sampler that are "not sampled" is a repeated theme, and we hadn't quite uncovered the relationship between "not sampled" and "is_threshold_reliable".
This PR solves the problem! There is still no such thing as "not sampled", instead we have "sampled with unknown randomness".💯
Co-authored-by: Kent Quirk <kentquirk@gmail.com>
|
This PR was marked stale due to lack of activity. It will be closed in 7 days. |
…ication into 4566_clarifications_for_Composite_Samplers_proposal
carlosalberto
changed the title
### Entities - Specify entity information via an environment variable. ([#4594](#4594)) ### Common - OTLP Exporters may allow devs to prepend a product identifier in `User-Agent` header. ([#4560](#4560)) -⚠️ **IMPORTANT**: Extending the set of standard attribute value types is no longer a breaking change. ([#4614](#4614)) ### OTEPs - Clarify in Composite Samplers OTEP the unreliable threshold case. ([#4569](#4569)) --------- Co-authored-by: Armin Ruech <7052238+arminru@users.noreply.github.com>
5 participants
Addressing issue #4566.
Fixes #4566
Changes
Clarifying the meaning of
IsAjustedCountReliable, providing guidance for selecting the Randomness value R for sampling decision making.