Closed lmolkova closed 7 months ago
It seems there were no discussions on this in https://github.com/open-telemetry/opentelemetry-specification/pull/2456 when it was introduced.
@laurit do you happen to remember if there was a specific reason to add a new attribute?
Also, graphql does not extend db
in yaml and does not put its attributes into db.
namespace. Is it intentional?
@laurit do you happen to remember if there was a specific reason to add a new attribute?
I believe I used the attribute names from the nodjs graphql instrumentation.
Also, graphql does not extend db in yaml and does not put its attributes into db. namespace. Is it intentional?
As far as I remember treating graphql as, or as a subset, of db was not considered during the review.
Well, I guess there we have some breaking changes :D
it's not clear to me the criteria for GraphQL being grouped under databases?
Honest question, why does it need to be under db.
?
Honest question, why does it need to be under
db.
?
If we don't consider GraphQL to be a database, it would totally be fine to not add db.
, but then we need to remove it from database folder (semantic-conventions/blob/main/docs/database/graphql.md)
and stop listing it under DB semconv https://github.com/open-telemetry/semantic-conventions/blob/44c830d3dabad9d59f340d92e774a1d5a2356743/docs/database/README.md?plain=1#L27
Both options work for me.
but then we need to remove it from database folder
Oh, I didn't know its location. Thanks.
I agree on moving it.
IMHO, it should have its own folder, since it doesn't seem to fit into any other existing one.
Should be covered by
db.operation
.