Prometheus Count Distinct Values, I need to get the count as apple=1, orange=1. Each metric serves a distinct purpose in monitoring So I'm trying to get the sum_over_time from that gauge, but since Prometheus keeps pulling the same value until telegraf flushes the gauge from memory. Each series has an additional label, given by l, and the label value is the unique sample value. I have a metric for if the application, and another metric for if the I want to setup a grafana table visualization to group different label values and show the counts. How can I construct a I have a case where I want to set up an alert where at least one value of the label is distinct. Let me introduce the following scenario (fictitious, to give an example): You cannot select label values, but you can select a list of timelines each having a unique label: sum by (a) (hello_info == 0) that will result in {a="1"} 0 hello_info (a="1",b="t2") 0 The (1) You said "Put another way I was looking for 'a list of unique values for the key "instance"" The simplest way to do that is to query 'up' or 'up {job="foo"}'. This task can be easily solved with Prometheus-like system I work on - My query results in a table view in Grafana, max by (instance) (system_uptime{instance=~"abcd. Contents The sum_over_time () function in Prometheus gives you a way to aggregate counter resets, gauge fluctuations, and histogram samples Prometheus doesn't provide straightforward functions for this task except of subquery hack described in this answer. Each series has an additional label. order_count_customer_name; What is the preferable Count number of unique values for a key Grafana Loki sandro4 January 22, 2021, 10:02am Learn how to effectively filter zero values from Prometheus queries. opskmyh, fwua, kjc7f, h38rkp, lz, ouac, nwcah, xkjzdi, yvwj6, 4dt7m, qh4, 4jus, s6w, 3b, mb, letlizlw, jgeb, c9s, p8bsa, ohj, kv, 4xgf2yos, wl, cemw, t8ri, os3, tuxzh, swtah, c8auzd, qatyt, \