Hello,
I'm opening this issue as a follow up to a discussion in open-telemetry/opentelemetry-proto#346 (comment) where it appeared unclear whether processors of telemetry data should be designed to support repeating keys in attribute lists.
I do not know what the right answer should be, but I couldn't find a definitive statement in the Attributes Specification to help guide design decisions that would depend on this information.
There is one mention of "unique keys" in the Attributes Limits, but it is unclear whether "unique" refers to keys being unique or the number of keys in the attribute lists:
- set a limit of unique attribute keys such that:
- for each unique attribute key, addition of which would result in exceeding the limit, SDK MUST discard that key/value pair.
Let me know if I can provide any more context that could help address this issue.
Hello,
I'm opening this issue as a follow up to a discussion in open-telemetry/opentelemetry-proto#346 (comment) where it appeared unclear whether processors of telemetry data should be designed to support repeating keys in attribute lists.
I do not know what the right answer should be, but I couldn't find a definitive statement in the Attributes Specification to help guide design decisions that would depend on this information.
There is one mention of "unique keys" in the Attributes Limits, but it is unclear whether "unique" refers to keys being unique or the number of keys in the attribute lists:
Let me know if I can provide any more context that could help address this issue.