Depends on
https://github.com/open-telemetry/opentelemetry-collector/pull/12856Resolves#12676
This is a reboot of #11311, incorporating metrics defined in the
[component telemetry
RFC](https://github.com/open-telemetry/opentelemetry-collector/blob/main/docs/rfcs/component-universal-telemetry.md)
and attributes added in #12617.
The basic pattern is:
- When building any pipeline component which produces data, wrap the
"next consumer" with instrumentation to measure the number of items
being passed. This wrapped consumer is then passed into the constructor
of the component.
- When building any pipeline component which consumes data, wrap the
component itself. This wrapped consumer is saved onto the graph node so
that it can be retrieved during graph assembly.
---------
Co-authored-by: Pablo Baeyens <pablo.baeyens@datadoghq.com>
<!--Ex. Fixing a bug - Describe the bug and how this fixes the issue.
Ex. Adding a feature - Explain what this achieves.-->
#### Description
Wraps the component ID and type into the component's start retuned
error. If `telemetry::logs::level` is < error (e.g fatal), the service
error message does not give information about which pipeline component
failed:
Current error message:
```
$ otelcontribcol --config config.yaml
Error: cannot start pipelines: start function error
2024/12/10 10:07:25 collector server run finished with error: cannot start pipelines: start function error
```
With these changes:
```
$ otelcontribcol --config config.yaml
Error: cannot start pipelines: failed to start geoip processor: start
function error
2024/12/10 09:58:29 collector server run finished with error: cannot
start pipelines: failed to start geoip processor: start function error
```
<!-- Issue number if applicable -->
#### Link to tracking issue
https://github.com/open-telemetry/opentelemetry-collector/issues/10426
<!--Describe what testing was performed and which tests were added.-->
#### Testing
<!--Describe the documentation added.-->
#### Documentation
<!--Please delete paragraphs that you did not use before submitting.-->
---------
Co-authored-by: Bogdan Drutu <bogdandrutu@gmail.com>
Co-authored-by: Evan Bradley <11745660+evan-bradley@users.noreply.github.com>
#### Description
In #10058 I mentioned:
> There is a tangentially related issue with PermanentErrors and the
underlying finite state machine that governs transitions between
statuses. Currently, a PermanentError is a final state. That is, once a
component enters this state, no further transitions are allowed. In
light of the work I did on the alternative health check extension, I
believe we should allow a transition from PermanentError to Stopping to
consistently prioritize lifecycle events for components. This transition
also make sense from a practical perspective. A component in a
PermanentError state is one that has been started and is running,
although in a likely degraded state. The collector will call shutdown on
the component (when the collector is shutting down) and we should allow
the status to reflect that.
This PR makes the suggested change and updates the documentation to
reflect that. As this is an internal change, I have not included a
changelog. Also note, we can close#10058 after this as we've already
removed status aggregation from core during the recent component status
refactor.
<!-- Issue number if applicable -->
#### Link to tracking issue
Fixes#10058
<!--Describe what testing was performed and which tests were added.-->
#### Testing
units
<!--Describe the documentation added.-->
#### Documentation
Updated docs/component-status.md and associated diagram.
<!--Please delete paragraphs that you did not use before submitting.-->
Co-authored-by: Tyler Helmuth <12352919+TylerHelmuth@users.noreply.github.com>
Co-authored-by: Antoine Toulme <atoulme@splunk.com>
#### Description
[whitespace](https://golangci-lint.run/usage/linters/#whitespace) is a
linter that checks for unnecessary newlines at the start and end of
functions.
Signed-off-by: Matthieu MOREL <matthieu.morel35@gmail.com>
#### Description
This PR removes the `Ready` method from `status.Reporter`. This check
was necessary when `ReportStatus` was on `TelemetrySettings` as it
should not be called until component start-time, but was available to a
component at create-time. This method now resides on `host` which is not
available until component start, making this check unnecessary. This is
an internal refactor that should not affect any other users. I'm not
sure if this needs a changelog, but I generated one anyways and set the
type to breaking, as it removes a method from an interface.
#### Link to tracking issue
Fixes#10792
#### Testing
Units
---------
Co-authored-by: Pablo Baeyens <pablo.baeyens@datadoghq.com>
Co-authored-by: Alex Boten <223565+codeboten@users.noreply.github.com>
This PR follows #11321 by splitting up the primary test file into a few
related topics. I believe this will make further refactoring PRs easier
to follow.
Daniel Jaglowski
Matthieu MOREL
Roger Coll
Matthew Wear