-
Notifications
You must be signed in to change notification settings - Fork 790
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
Rename two more modules / maven artifactIds? #668
Comments
How about moving support for |
The naming does seem confusing, especially once it's out of beta, our artifact name Merging |
At this point at least,
Can you open an issue to track this? I don't think I've run into this.
I'm mixed on this. Both And they both share On the other hand Ok I'm not mixed on this anymore, I agree with re-alignment 😄. What do you think of
I'd also be happy with just combining |
It is always nice to read Trask’s train of thoughts :) I don’t think separate module just for 1 annotation is justified. |
Ah I guess I got it because I use the API just to get the current span ID for some debugging. Dunno how representative of usage this is, but I'd be surprised if long term we can get away with users not using the API at all - one of the first things I'd do in the app is get the current span and stuff tags in for whether Though that's a lot of context (not that one, that's coming up :P), but in more practical terms, the better question would have been do we expect any repackaged agent to function properly without this? I think the potential for manual API usage will exist for all agents and they need this, whereas they may decide to exclude some of the library instrumentations. So if it seems required to create any sort of OTel agent anyways, there doesn't seem to be that much value in having a separate artifact. I also notice
Messed was the wrong word I think, more just confusing, was seeing packages I didn't expect in the debugger. Don't think this can be avoided when doing shading
I think it's a lot of logic, just to process 1 measly annotation ;) But personally agree that combining and possibly splitting out later is fine, there's less worry about refactoring agent modules than there is if this was manual instrumentation. |
Yeah, this instrumentation needs to be reviewed, I did not have a good grasp of context propagation when I wrote it.
Only if OpenTelemetry API is also on the classpath, due to https://github.com/open-telemetry/opentelemetry-java-instrumentation/blob/master/instrumentation/opentelemetry-api-beta/src/main/java/io/opentelemetry/auto/instrumentation/opentelemetryapi/GrpcContextInstrumentation.java#L64-L65
I think the same could be said for There's benefit to it still being a separate module (e.g. |
I didn't know about the benefits :P But yeah the suggestion was mainly about the published artifacts, and indeed it probably applies to other modules too. |
Ok, based on discussion, a concrete proposal:
I don't have much opinion about whether to publish these "critical instrumentation" artifacts ( One possible advantage to publishing them separately is in case a vendor wants to override/fork/patch one of them, but I'm thinking that's probably not a real use case. |
LGTM - took a second look and just realized we support all these different annotations that's pretty cool. Might rename it to |
Module name is currently
Done in #929 |
The
opentelemetry-api-beta
artifactId is different from other instrumentation, which all start withopentelemetry-auto-*
, and theopentelemetry-auto-annotations
artifactId is different, because it has no version:What do you think about renaming the
opentelemetry-api-beta
module tootel-api-beta
, and applying the normal artifactId prefix, which would make the artifactIdopentelemetry-auto-otel-api-beta
?Or keeping the same module name, but still applying the normal artifact prefix, which would make the artifactId
opentelemetry-auto-opentelemetry-api-beta
?For
opentelemetry-auto-annotations
, it serves two purposes, one is for otel annotations, and one is for other annotations. What do you think of versioning it based on the otel annotation version that it supports? So for now, making itannotations-beta
(similar tootel-api-beta
) and then moving to real version numbers once we hit1.0
.Or, another option is to split up the annotations module:
:instrumentation:annotations:annotations-otel-beta
(beta
to be replaced by1.0
in GA):instrumentation:annotations:annotations-other
(with no version, as it targets several different annotations from different libraries):instrumentation:annotations:annotations-common
(we have some configuration parsing code that is shared)The text was updated successfully, but these errors were encountered: