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
[Feature]: add support for "Service Performance Monitoring (SPM)" #440
Comments
I was trying to add the following values, and I can see that Prometheus scrapes the metrics from the controller and query, but the monitor is still not showing any metrics.
|
Hi @maozza , you will need to deploy the opentelemetry-collector helm chart to achieve this 🙂 The collector takes the spans and produces prometheus metrics from it, if you add the spanmetrics to its pipelines🙂 In my specific deployment I could even replace the jaeger collector completely with the open telemetry. I would be happy to add a respective PR, however a question to the maintainers before doing so: Is this generally a direction you want to go into? 🙂 I can see that an argument can be made to just disable the collector in this chart and deploy the open telemetry one with a separate one. The questions I feel is if you see the case of using the open telemetry collector as the standard one in the future or the jaeger one. Just wanted to figure before I start coding 😉 |
As I understand it, https://opentelemetry.io/docs/collector/ OTEL collector does not replace Jaeger as a storage/querying piece of tracing data. It receives/processes/exports tracing data. OTEL collector would still need a backend to send traces to and that could be Jaeger (jaeger can act as an exporter for OTEL collector). So i don't see how you would disable jaeger collector and make the collection of traces and storage of them work end to end. EDIT: |
Thanks for your input @mehta-ankit 🙂 No, it indeed does not replace storage and querying. However, it can use one of its exporters to write directly to the storage backend, e.g. elastic search, from which query could retrieve the traces then 🙂 Right now my setup to enable SPM is as well jaeger-agent -> open telemetry collector -> jaeger-collector -> elastic search 🙂 So if supporting SPM out of the box, I would suggest a similar setup for this chart as it alters it the least 🙂 I just wanted to confirm that the collector to collector approach would be the right one to take for a PR 🙂 |
I would appreciate adding this as well. I investigated how to set this up and found the Getting Started instructions in the Jaeger SPM docs much too terse and just boiling down to "just use the all-in-one". |
@FieteO Have you managed to complete this setup? Because I'm in this very same situation right now. |
@ezradibiase Unfortunately not. I did not pursue it further though. |
Requirement
I want to enable Service Performance Monitoring (SPM) feature with helm installation
install procedure:
Problem
will it be possible to add support for Service Performance Monitoring (SPM) in the helm chart?
more info can be found here -> https://www.jaegertracing.io/docs/1.41/spm/
Proposal
No response
Open questions
No response
The text was updated successfully, but these errors were encountered: