the documentation of grafana alloy clustering seems focus on scraping metrics. but I would like to find out how does it work when metrics have to be pushed into Alloy with clustering enabled ? does it have load balancing mechanism etc. ? thanks.
It doesn’t. For metrics that are pushed to Alloy (or logs, for that matter) you’d put a load balancer in front of your alloy agents, and whether they are clustered has no effect.
You can see the list of scraping components that support auto-distribution (clustering) here: Clustering | Grafana Alloy documentation