Hi there, currently I am trying signoz but I struggle with the scraping of kubernetes /metrics endpoints. I searched here, but I don't understand the solutions posted here. I am using signoz in a talos kubernetes cluster. First thing I figured out was, that I had to install the k8s-infra helm chart next to the signoz helm chart to get logs and metrics from my cluster. Something that I think was not documented. Now I want that signoz collects the metrics from my pods. I found this in the internet:

podAnnotations:
        <http://signoz.io/scrape|signoz.io/scrape>: "true"
        <http://signoz.io/port|signoz.io/port>: "6060"
        <http://signoz.io/path|signoz.io/path>: "/metrics"

but it doesn't work. What must be done that signoz scrapes metrics via pod annotations? I don't want to - or it's not practical - to define a block in some config for every endpoint I want to scrape.

Some pointers here https://github.com/SigNoz/charts/issues/445#issue-2333727471

Oh, thank you so much! I just wonder - if I selfhost signoz, is the k8s-infra required?

signoz chart runs the main signoz installation. k8s-infra chart for collecting and sending logs/metrics/traces data from k8s cluster to main signoz installation.

ok, then there is an issues in the docs, as they tell the helm chart will install it, but they dont

would you mind creating an issue here https://github.com/SigNoz/signoz-web?

will do so - but then, if i have already signoz running in the cluster - will this be sufficient to collect prometheus metrics, or is it only possible via the infra-k8s chart?

What I shared is applicable for k8s-infra. There is addl collector (signoz-collector-metrics, please note only one instance signoz-collector-metrics exist and it should be one to avoid duplication) that runs as part of SigNoz installation to help for the same. You can override it's config to get metrics scraped.