First of all, note that alerting rules live on your Prometheus
server. Alert routing and grouping happens on the Alertmanager.
To strictly answer your question: You control how often an alerting
rule is evaluated in the same way as any rule. For example, you can
set it for each rule group, see
https://prometheus.io/docs/prometheus/latest/configuration/recording_rules/#rule_group
However, I don't think you want to sat that to 1d. Prometheus rules,
including alerting rules, just create other time series, and you
usually want those to get a new sample at least every 2m.
Perhaps the group_interval or the repeat_interval in the Alertmanager
config is more what you want, but I'm not sure. I recommend to read up
again in the Alertmanager documentation what they do.
In any case, you can set the group_interval and the repeat_interval
individually for every part of your routing tree, see
https://prometheus.io/docs/alerting/latest/configuration/#route
--
Björn Rabenstein
[PGP-ID] 0x851C3DA17D748D03
[email]
bjo...@rabenste.in