Open rus-99-pk opened 6 months ago
This comes from the old builtinMetricMaps
in cmd/postgres_exporter/postgres_exporter.go
. This is on the TODO list to be migrated to the collector
package. This will enable it to be
Due to the large number of metrics, the third-party application that receives these metrics freezes.
This sounds like a bug, Prometheus should not "freeze" when ingesting data. There are also scrape sample limits in the configuration to help avoid problems like this.
@SuperQ, thank you!
This sounds like a bug, Prometheus should not "freeze" when ingesting data. There are also scrape sample limits in the configuration to help avoid problems like this.
Yes, Prometheus doesn't freeze. I talked about my third-party application, that receives these metrics from Prometheus.
This comes from the old builtinMetricMaps in cmd/postgres_exporter/postgres_exporter.go. This is on the TODO list to be migrated to the collector package. This will enable it to be
I understand correctly, that it will realized in the future, or not? If is it right, than when will it happen?
Hello!
Proposal
I suggest adding functions to exclude some metrics in postgres_exporter.
Because I have a large number of sessions and queries in the database, so this number is showed in the form of metrics. Due to the large number of metrics, the third-party application that receives these metrics freezes.
To avoid this, it is necessary to disable a number of metrics, for example pg_stat_activity_count.
I think it will be useful, because i didn't find this in standart functionality.