Example Queries

Video Lecture

Description

Now that we have at least 2 scape targets, we can begin to run some more interesting queries that involve multiple scrape targets.

scrape_duration_seconds

This shows 2 time series results. 1 for each target,

we can filter for 1 target by including either the instance, or job labels

scrape_duration_seconds{instance="localhost:9100"}

Regular Expressions

We can also use regular expressions. Go back to the console view and query for node_cpu_seconds_total We should get about 8 time series results. Lets filter for everything with mode containing irq

node_cpu_seconds_total{mode=~".*irq"}

All regular expressions in Prometheus use the RE2 syntax.

Data Types

Instant vector

A set of time series containing a single sample for each time series, all sharing the same timestamp `scrape_duration_seconds{instance="localhost:9100"}

Range Vector

A set of time series containing a range of data points over time for each time series.

Return a whole range of scrape_duration_seconds (in this case 5 minutes) for the same vector, making it a range vector.

node_netstat_Tcp_InSegs{instance="localhost:9100"}[5m] node_netstat_Tcp_InSegs{instance="localhost:9100"}[1m] node_netstat_Tcp_InSegs{instance="localhost:9100"}[30s]

Press the graph, we get an error saying that expression type "range vector" for range query, must be Scalar or instant Vector

So when in the graph view, remove the [5m] range option.

Functions

rate(scrape_duration_seconds{instance="localhost:9100"}[1m:20s])

Start with node_netstat_Tcp_InSegs

create a range vector covering 1m node_netstat_Tcp_InSegs[10m]

filter for the job prometheus node_netstat_Tcp_InSegs{job="prometheus"}[10m]

calculates the per-second average rate of increase of the time series in the range vector rate(node_netstat_Tcp_InSegs[10m])

sum the 2 values sum(go_threads)

Sub Queries

Start with this instant vector node_netstat_Tcp_InSegs

Convert ot to a Range Vecorer and then convert it back to an instant vector using rate rate(node_netstat_Tcp_InSegs[1m])

Wrap it in the ceikling funtion ceil(rate(node_netstat_Tcp_InSegs[1m]))

Convert it to a range vecort and get the per-second derivative of the time series deriv(ceil(rate(node_netstat_Tcp_InSegs[1m]))[1m:])

https://prometheus.io/docs/prometheus/latest/querying/functions/

https://github.com/prometheus/prometheus/wiki/Default-port-allocations