I don't think this is currently possible in prometheus. If you are using grafana though, you can realize this using variables.

Try something like this:
mysql_global_status_wsrep_cluster_info unless ignoring(__dst__) (label_replace(mysql_global_status_wsrep_cluster_info, "__dst__", "$1", "cluster-uuid", "(.*)") == label_replace(mysql_global_status_wsrep_cluster_info, "__dst__", "$1", "local-uuid", "(.*)"))

Hi @IldarMinaev, your answer was very useful. Thanks.

Now I need to create a chart from this result, something like this:

mysql_global_status_wsrep_cluster_info[1h:1d] unless ignoring(__dst__) (label_replace(mysql_global_status_wsrep_cluster_info, "__dst__", "$1", "cluster-uuid", "(.*)") == label_replace(mysql_global_status_wsrep_cluster_info, "__dst__", "$1", "local-uuid", "(.*)"))

But I'm getting this error:
Error executing query: 1:1: parse error: binary expression must contain only scalar and instant vector types

I also tried something like this:
(mysql_global_status_wsrep_cluster_info unless ignoring(__dst__) (label_replace(mysql_global_status_wsrep_cluster_info, "__dst__", "$1", "cluster-uuid", "(.*)") == label_replace(mysql_global_status_wsrep_cluster_info, "__dst__", "$1", "local-uuid", "(.*)")))[1h]

But I got this: Error executing query: 1:227: parse error: ranges only allowed for vector selectors

Hello from the bug scrub.

Apologies for the delay in responding to this.

This is more of a usage question and so converting to a discussion.

When comparing label values for a single metric in Prometheus, you are essentially looking at the different label combinations that identify specific instances or dimensions of that metric. Prometheus uses labels to provide additional identifying information about the data being collected, allowing you to filter and aggregate metrics in powerful ways.