*Empty activeAlertmanagers list after executing reload method (0.001...5 sec)*
1 Run prometheus with params (go run cmd/prometheus/main.go --config.file=config.yaml --web.enable-lifecycle) config.file: YAML 1 2 3 4 5 6 alerting: alertmanagers: - http_sd_configs: - refresh_interval: 20s url: http://xxx timeout: 1m 2 make request curl -sXPOST http://localhost:9090/-/reload 3 and then you can try to open http://localhost:9090/api/v1/alertmanagers 4 Output will be empty. looks like: { "status": "success", "data": { "activeAlertmanagers": [], "droppedAlertmanagers": [] } } 5 Only after executing peace of code prometheus will update the content of config refs: https://gitlab.ozon.ru/sre/prometheus/-/blob/9619d3fd3b05122d2fad9d80391b100f976d3764/notifier/notifier.go#L309 ^ https://gitlab.ozon.ru/sre/prometheus/-/blob/432005826d06e6d6ac8f2d3836f7fdf66d9c3092/discovery/legacymanager/manager.go#L243 ^ https://gitlab.ozon.ru/sre/prometheus/-/blob/432005826d06e6d6ac8f2d3836f7fdf66d9c3092/discovery/legacymanager/manager.go#L97 as a result, prometheus has a time window from 0.0001 to 5 seconds when downtime is possible. How do you cope with such a task? -- 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 prometheus-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/prometheus-users/4b4ba5e1-e261-455f-b842-345916f2b76fn%40googlegroups.com.
