Различия

Показаны различия между двумя версиями страницы.

--- service:grafana [19.02.2021 15:59] – создано viacheslav
+++ service:grafana [17.08.2024 12:43] (текущий) – [Некоторые метрики] viacheslav
@@ Строка 1: / Строка 1: @@
+====== Grafana ======
+https://grafana.com\\
+https://hub.docker.com/r/grafana/grafana/\\
+https://grafana.com/docs/grafana/latest/administration/configure-docker/
+Конфиг - /etc/grafana/grafana.ini
+===== Некоторые метрики =====
+<code bash>
+# CPU
+node_load1
+node_load5
+node_load15
+# CPU в процентах
+- (avg(irate(node_cpu_seconds_total{mode="idle"}[30m])) * 100)
+# RAM (MiB)
+node_memory_MemFree_bytes / 1024 / 1024
+node_memory_MemTotal_bytes / 1024 / 1024
+# HDD (GiB)
+node_filesystem_free_bytes{mountpoint="/"} / 1024 / 1024 / 1024
+node_filesystem_size_bytes{mountpoint="/"} / 1024 / 1024 / 1024
+# Disk usage (GiB)
+node_disk_read_bytes_total{device="sda"} / 1024 / 1024 / 1024
+node_disk_written_bytes_total{device="sda"} / 1024 / 1024 / 1024
+# System (from docker metrics)
+engine_daemon_engine_info
+# Legend: {{os}} {{architecture}} ({{kernel}})
+# Docker {{version}}
+# Boot time (Datetime default)
+node_boot_time_seconds * 1000
+# Uptime (Duration d hh:mm:ss)
+node_time_seconds - node_boot_time_seconds
+# Pending updates
+apt_upgrades_pending
+# Reboot required (+ value mapping)
+node_reboot_required
+# Services
+count(node_systemd_unit_state{state="active"}!=0)
+count(node_systemd_unit_state{state="inactive"}!=0)
+count(node_systemd_unit_state{state="failed"}!=0)
+# Docker containers status
+engine_daemon_container_states_containers{state="running"}
+engine_daemon_container_states_containers{state="stopped"}
+engine_daemon_container_states_containers{state="paused"}
+# Network
+node_network_receive_bytes_total{device="eth0"} / 1024 / 1024
+node_network_transmit_bytes_total{device="eth0"} / 1024 / 1024
+# Текущий трафик по всем интерфейсам (включая Докер, loopback и т. д.)
+sum(irate(node_network_receive_bytes_total[1m]))
+sum(irate(node_network_transmit_bytes_total[1m]))
+# Более аккуратный вариант
+sum(irate(node_network_receive_bytes_total{device=~"(wlp1s0|enp3s0)"}[1m]))
+sum(irate(node_network_transmit_bytes_total{device=~"(wlp1s0|enp3s0)"}[1m]))
+# Выцепить несколько имён метрик
+{__name__=~"metrics_containers_.*_cpu_percent"}
+#metrics_containers_frigate_cpu_percent{instance="glances:9091", job="glances", src="glances"}	7.388715365239294
+#metrics_containers_glances_cpu_percent{instance="glances:9091", job="glances", src="glances"}	0.6252216748768473
+#metrics_containers_nc_cpu_percent{instance="glances:9091", job="glances", src="glances"}	0.0629145728643216
+</code>
+===== Настройка за реверс-прокси =====
+Мало настроить реверс-прокси, нужно Графане рассказать, как работать.
+<code>
+[server]
+domain = example.com
+root_url = %(protocol)s://%(domain)s:%(http_port)s/grafana/
+serve_from_sub_path = true
+</code>
+https://grafana.com/tutorials/run-grafana-behind-a-proxy/#1
+===== Общий доступ к панели =====
+[[https://stackoverflow.com/questions/33111835/how-to-set-up-grafana-so-that-no-password-is-necessary-to-view-dashboards|How to set up Grafana so that no password is necessary to view dashboards]]
+https://grafana.com/docs/grafana/v7.5/auth/grafana/#anonymous-authentication
+===== Показать значения ярлыков внутри метрики =====
+К примеру, есть метрика вида
+  node_uname_info{domainname="(none)", instance="172.17.0.1:9100", job="host", machine="x86_64", nodename="user-xubuntuVM", release="5.4.0-65-generic", sysname="Linux", version="#73-Ubuntu SMP Mon Jan 18 17:25:17 UTC 2021"} 1
+Из неё нужно вытащить значение sysname и release.
+{{:service:pasted:20210225-094434.png}}
+Затем в правой колонке настроек выбрать
+  * Visualization -> Stat
+  * Display -> Text mode -> Name
+  * Display -> Graph mode -> None
+https://stackoverflow.com/questions/38525891/how-do-i-write-a-prometheus-query-that-returns-the-value-of-a-label