Changeset View
Changeset View
Standalone View
Standalone View
data/common/common.yaml
| Show First 20 Lines • Show All 3,157 Lines • ▼ Show 20 Lines | prometheus::server::config::global: | ||||
| external_labels: | external_labels: | ||||
| tenant: "%{lookup('thanos::tenant')}" | tenant: "%{lookup('thanos::tenant')}" | ||||
| replica: "%{lookup('thanos::replica')}" | replica: "%{lookup('thanos::replica')}" | ||||
| prometheus::server::config::static_scrape_configs: [] | prometheus::server::config::static_scrape_configs: [] | ||||
| prometheus::server::listen_network: "%{lookup('internal_network')}" | prometheus::server::listen_network: "%{lookup('internal_network')}" | ||||
| prometheus::server::listen_port: 9090 | prometheus::server::listen_port: 9090 | ||||
| prometheus::server::certname: pergamon.softwareheritage.org | prometheus::server::certname: pergamon.softwareheritage.org | ||||
vsellier: I suppose it's not the thanos query server defined here to avoid broadcasting the queries to… | |||||
Done Inline ActionsMy idea is to always contact the local prometheus server for alerting; When we'll set up alertmanager, I assume we'll want to do something similar (and only use thanos for aggregation and cross-tenant alerting) olasd: My idea is to always contact the local prometheus server for alerting; When we'll set up… | |||||
| prometheus::server::fqdn: pergamon.internal.softwareheritage.org | |||||
| swh::deploy::environment: production | swh::deploy::environment: production | ||||
| prometheus::static_labels: | prometheus::static_labels: | ||||
| instance: "%{::swh_hostname.internal_fqdn}" | instance: "%{::swh_hostname.internal_fqdn}" | ||||
| environment: "%{lookup('swh::deploy::environment')}" | environment: "%{lookup('swh::deploy::environment')}" | ||||
| prometheus::node::listen_network: "%{lookup('internal_network')}" | prometheus::node::listen_network: "%{lookup('internal_network')}" | ||||
| ▲ Show 20 Lines • Show All 805 Lines • Show Last 20 Lines | |||||
I suppose it's not the thanos query server defined here to avoid broadcasting the queries to all the prometheus (and because it's live data) ?