max / makenotwork

2.1 KB · 70 lines History Blame Raw

1	//! offline alert/recovery messages. Domain half of the Alerter split;
2	//! shared dispatch/cooldown plumbing lives in the parent module.
3
4	use tracing::instrument;
5
6	use super::{AlertCategory, AlertMeta, Alerter};
7
8	impl Alerter {
9	/// All monitored targets are unreachable, likely a network issue with PoM itself.
10	#[instrument(skip_all)]
11	pub async fn send_monitoring_offline_alert(&self, target_count: usize) {
12	let alert_key = "monitoring:self";
13	let subject = format!("[PoM] all {target_count} targets unreachable");
14	let body = format!(
15	"All {target_count} monitored targets are non-operational.\n\
16	This likely indicates a network issue with the PoM instance itself,\n\
17	not an actual outage of all targets.\n\n\
18	Instance: {}\n\
19	Time: {}\n\n\
20	- PoM",
21	self.instance_name,
22	chrono::Utc::now().to_rfc3339(),
23	);
24
25	self.fire_failure(
26	&subject,
27	&body,
28	"critical",
29	"pom-monitoring",
30	Some("self"),
31	AlertMeta {
32	key: alert_key,
33	category: AlertCategory::MonitoringOffline,
34	from: None,
35	to: None,
36	error: None,
37	},
38	)
39	.await;
40	}
41
42	/// At least one target is reachable again after a monitoring-offline event.
43	#[instrument(skip_all)]
44	pub async fn send_monitoring_recovery(&self) {
45	let alert_key = "monitoring:self";
46	let subject = "[PoM] monitoring recovered".to_string();
47	let body = format!(
48	"At least one target is reachable again.\n\
49	Instance: {}\n\
50	Time: {}\n\n\
51	- PoM",
52	self.instance_name,
53	chrono::Utc::now().to_rfc3339(),
54	);
55
56	self.fire_recovery(
57	&subject,
58	&body,
59	AlertMeta {
60	key: alert_key,
61	category: AlertCategory::MonitoringRecovery,
62	from: None,
63	to: None,
64	error: None,
65	},
66	)
67	.await;
68	}
69	}
70