Loki/Promtail: different label set on identically configured k8s setup

I have a test kubernetes cluster where I practiced with loki/promtail. After some months I created a clone for production and I notice a totally different set of labels in loki:

  • The configuration are exactly identical (yaml file installed with loki-stack)

  • Releases of grafana/loki/promtail are different

being pretty new to loki I thought it was promtail that added the label such as instance or name and I tried to use same promtail version, to no avail.

before trying any possible release, I’d like to have some advice on who’s responsible for adding these labels.


I got to the conclusion that the culprit is promtail that does not add labels that are set on the pod/service.

I guess the reason is a different configuration due to a big change in the helm chart (loki-stack that in turn depends on promtail).

Does anybody know how to enable those settings (default in previous chart) in the new helm chart?