Blog post regarding ECS and OTel SemConv Convergence

[reyang]

Note: we would like the OTel blog published 7am Amsterdam time on April 18th which is 10pm PT on April 17th.

Background context

The stakeholders of ECS (Elastic Common Schema) and OpenTelemetry have been making progress since the past 12 months with the outcome of OTEP 199. Since the OTEP 199 got approved/merged, we've had multiple discussions among the ECS stakeholders from Elastic, OpenTelemetry GC (Governance Committee) and TC (Technical Committee) members, now we've reached to a point that we are ready to publicly announce this merge in the upcoming EU KubeCon event ❤️🍾🥳!

What's the communication plan

The plan is that Elastic and OpenTelemetry will each publish an official post/announcement at 7am Amsterdam time on April 18th which is 10pm PT on April 17th (right before KubeCon). This PR covers the blog post from OpenTelemetry, while Elastic will publish their announcement to https://elastic.co/blog/ecs-elastic-common-schema-otel-opentelemetry-announcement (note that the link will return 404 at this moment, and will become available at the publishing time).

Additional info

A draft Google Doc was prepared and went through several rounds of internal reviews in the past two weeks, which resulted in this PR after the alignment. The original doc can be found from https://docs.google.com/document/d/1ymr4sVG-Npdr1-hys-PKhF6pZXc_bNYmR1fee_n9GWM/edit?usp=sharing.

@AlexanderWert @open-telemetry/governance-committee @open-telemetry/technical-committee

[chalin]

FYI, I rebased, ran the formatter and updated the refcache. I'll soon submit a few suggested copyedits.

[chalin]

Copyedits and a fix for the build warning. See inline comments.

[cartermp]

I think we should signal in this post that this means breaking changes are afoot (and they're significant). There's a lot of people who have been using http instrumentation that's been stable for a long time, and all of that is changing. It will hopefully be the case that all the telemetry backends can make this transition seamless for their users, but we should also call that out as a goal and that we're working with backend vendors to define that transition path.

[reyang]

a lot of people who have been using http instrumentation that's been stable for a long time, and all of that is changing

Where is the "stable for a long time" coming from?

It will hopefully be the case that all the telemetry backends can make this transition seamless for their users, but we should also call that out as a goal and that we're working with backend vendors to define that transition path.

This is covered by the OTEP 199.

[cartermp]

HTTP instrumentions haven't changed nearly close to this extent in a long, long time. I've been able to query on http.status_code and http.method since...I'd like to say Fall of 20201?...and run those same queries. All of that is now changing. That's a massive breaking change. I don't think it's sufficient to just link to an OTEP. We can do it without doom and gloom, but I think that a change of this scale should be explicitly called out with a summarization of what the current thinking is (i.e., end-users don't need to act but telemetry backends do) from the very start.

[austinlparker]

I'll take responsibility for merging/publishing this fyi.

[austinlparker]

@cartermp FWIW I don't think we'll be able to make edits on this post as it seems that this language was hashed out w/Elastic. I do think our path forward here should be to publish a companion piece/explainer and link to it.

[alolita]

Can we also reference the other related PRs in the References section of the post.

For example,
open-telemetry/oteps#197 (which is the original proposal)

language pre-decided upon

[cartermp]

I do think our path forward here should be to publish a companion piece/explainer and link to it.

Okay, that's fair. I'll engage with that when we do it.

[reyang]

Can we also reference the other related PRs in the References section of the post.

For example, open-telemetry/oteps#197 (which is the original proposal)

@alolita Thanks for the suggestion! I've added more reference links, PTAL.

[austinlparker]

Need to get checks green before merge but I'm OK with copy as-is.

[reyang]

Need to get checks green before merge but I'm OK with copy as-is.

I've turned all the CI to green.

[austinlparker]

Goin' live!