-
Notifications
You must be signed in to change notification settings - Fork 2.4k
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
Prometheus scrape compatibility #9984
Comments
@dashpole @Aneurysm9 any thoughts on priority for this one? |
I think this is important. See #4980. We had said at a SIG meeting (#5080 (comment)) that it would be possible to do this using config sources. Is that possible now, and this is just a documentation issue? Or does config sources not solve the problem of having to escape $? |
@dashpole what are all the examples in prometheus config where you use
Because the expand PR is not approved, and because of backwards compatible reason |
The examples are almost entirely inside the My original thinking in #4980 was that being able to reference a separate file would do the trick. It would also enable other uses of the collector, such as using it with the Prometheus operator, where the operator expects to manage the Prometheus config file itself. I think we would lose the config "watching" functionality if we did that, though. |
I would suggest a different path. Implement a "MapConverter" (currently named |
Ooh, fancy. That works for me |
@dashpole if we are fine with this path, we should start by setting up the directory where we will have map converters in the contrib (keep in mind that we will have in the future providers as well). see https://github.com/open-telemetry/opentelemetry-collector/tree/main/config |
@dashpole another idea is to omit |
This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping |
Any updates here? |
How widespread is the actual impact of users relying on $1 as an environment variable to work? Seems like that would be extremely rare at best. |
Its relatively common when using relabel rules. https://github.com/prometheus/prometheus/blob/main/documentation/examples/prometheus-kubernetes.yml is the recommended config for prometheus on k8s, for example, and contains a few $1's in examples. |
@dashpole I think there's a misunderstanding. I am talking about |
Ah, yes. I did misunderstand your question. I've never seen anyone use $1 as an environment variable. |
I found another incompatibility as well. In Prometheus server, the relabel configuration applies an implicit "any" match |
Thank you for reporting this @otterley |
@bogdandrutu can we remove the stale label and have a convo on how we plan to address this please? |
Please bring this issue for discussion at the next SIG meeting? |
This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping |
This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping |
not stale |
This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping |
not stale |
This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping |
This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping |
Thankful to find this GH issue after wasting a solid half a day trying to figure out why my relabel configs weren't working. As an experienced Prometheus user starting to integrate the OTEL collector into our infrastructure this is a significantly confusing footgun. |
This issue stumped me for a solid week. Not having Yaml config interoperability for relabel configs is the mind killer. I am tired. |
I think open-telemetry/opentelemetry-collector/issues/9531 would help here |
Is your feature request related to a problem? Please describe.
Currently, when writing
relabel_configs
in the collector config prometheusreceiver one can not use$
but has to use$$
.While I understand why this is the case today, I see that folks coming from Prometheus and wanting to use the collector as a drop-in replacement, run into this UX issue, for example most recently aws-observability/aws-otel-collector#1201
Describe the solution you'd like
Make scrape configs 100% Prometheus compatible.
The text was updated successfully, but these errors were encountered: