Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Clarify the interpretation of
SpanKind
#337Clarify the interpretation of
SpanKind
#337Changes from 5 commits
c7ff294
6f30bff
dfc3cc9
1b36fbc
e03974d
7595543
File filter
Filter by extension
Conversations
Jump to
There are no files selected for viewing
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Some RPC frameworks (well, not really 'P' as in 'procedure) support one-way calls, that may or may not be acked on delivery, but certainly do not provide for a result to be returned to the caller (e.g. firing off a UDP message).
What span_kind should the caller use in that case? Producer?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think this question points out that
SpanKind
is still flawed.The question is whether these two should be independent or not. Can you be a CLIENT/PRODUCER? or a SERVER/CONSUMER? I.e., can you declare that there is a remote parent/child and an asynchronous relationship?
Can you be an internal span that has a PRODUCER/CONSUMER relationship with its child? These are capabilities we found in OpenTracing via ChildOf and FollowsFrom.
I am still skeptical of this field, but at least in the current state of this PR my understanding of the intention is closer to the group's.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@SergeyKanzhelev what do you think?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I agree with you that this PR improves current spec, but not addresses the problem of SpanKind expressiveness. Similar to #65 we just stalling on decision as not many consumer uses this information to the full extent.
For me - important thing is to distinguish incoming, internal and outgoing calls. As we had in OpenCensus. Mostly for the in-proc aggregation and analysis.