On Sun, 2 Jan 2022 at 20:43, Brian Candler <[email protected]> wrote: [...] > A subquery doesn't cut it here, because it resamples the data. The subquery > "(foo == N)[2d:1s]" gives an approximation, but for a given point you'll see > multiple points at 1 second intervals (until the time where foo != N)
If you don't mind the resampling then you can usefully make use of the experimental '@' modifier[1] to avoid needing to query for N first: max_over_time(foo[2d] @ end()) == foo For example, that is easily graphable like so: https://prometheus.demo.do.prometheus.io/graph?g0.expr=max_over_time(node_filefd_allocated%5B1d%5D%20%40%20end())%20%3D%3D%20node_filefd_allocated&g0.tab=0&g0.stacked=0&g0.range_input=1d&g0.step_input=10 (But note I had to put the step at 10s; the range vector can find values that your step may otherwise skip over.) [1]: https://prometheus.io/docs/prometheus/latest/querying/basics/#modifier -- You received this message because you are subscribed to the Google Groups "Prometheus Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/prometheus-users/CAP9KPhAfHxSLCSQrNorOkymoLpswpfXL1wT5zg8kmfaeRuGtvA%40mail.gmail.com.
