Files
infra/ansible/roles/monitoring_server/templates/alertmanager.yml.j2
T
DmitryandClaude Opus 5 c676be81ec Capture current Ansible control plane state
Commit the accumulated infrastructure work that was living only in the
working tree: monitoring stack, emergency access/bot, gyro allocator,
grimmory, adguard, backup audit and the OpenCode agent definitions.

Also ignore Python bytecode, local archives and Nix/direnv artifacts.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01GTocXkGUUazHdKKd3r9k71
2026-08-26 21:39:28 +03:00

41 lines
1.2 KiB
Django/Jinja

global:
resolve_timeout: 5m
route:
receiver: telegram-warning
group_by: [alertname, instance]
group_wait: 30s
group_interval: 10m
repeat_interval: 6h
routes:
- matchers:
- severity="critical"
receiver: telegram-critical
repeat_interval: 2h
inhibit_rules:
- source_matchers:
- alertname="ExporterDown"
target_matchers:
- alertname="LanServiceUnavailable"
equal: [instance]
receivers:
- name: telegram-warning
telegram_configs:
- bot_token: "{{ monitoring_telegram_bot_token }}"
chat_id: {{ monitoring_telegram_chat_id }}
send_resolved: true
message: |-
[{{ "{{" }} .Status | toUpper {{ "}}" }}] {{ "{{" }} .CommonLabels.alertname {{ "}}" }}
{{ "{{" }} .CommonAnnotations.summary {{ "}}" }}
- name: telegram-critical
telegram_configs:
- bot_token: "{{ monitoring_telegram_bot_token }}"
chat_id: {{ monitoring_telegram_chat_id }}
send_resolved: true
message: |-
[{{ "{{" }} .Status | toUpper {{ "}}" }}] CRITICAL: {{ "{{" }} .CommonLabels.alertname {{ "}}" }}
{{ "{{" }} .CommonAnnotations.summary {{ "}}" }}