import{_ as o,o as s,c as e,R as n}from"./chunks/framework.1625126e.js";const F=JSON.parse('{"title":"","description":"","frontmatter":{},"headers":[],"relativePath":"en_US/guide/notifications.md","filePath":"en_US/guide/notifications.md","lastUpdated":1691117191000}'),t={name:"en_US/guide/notifications.md"},a=n(`
Nezha Monitoring supports monitoring of server load, CPU, memory, hard disk, data transfer, monthly data transfer, number of processes, number of connections, and sends alarm notifications when one of these items reaches a user-set limit.
#NEZHA# is the panel message placeholder, the panel will automatically replace the placeholder with the actual message when it triggers the notification
The content of Body is in JSON format：When the request type is FORM，the value is in the form of key:value，value can contain placeholders that will be automatically replaced when notified. When the request type is JSON It will only do string substitution and submit to the URL directly.
Placeholders can also be placed inside the URL, and it will perform a simple string substitution when requested.
Refer to the example below, it is very flexible.
Bark Example
Name: Bark
URL: The first part is the key, followed by three matches/: key/: body or/: key/: title/: body or/: key/: category/: title/: body
Notes for this method：The XXXXXX in botXXXXXX is the token provided when you follow the official @Botfather in Telegram and enter /newbot to create a new bot. (In the line after Use this token to access the HTTP API). The 'bot' are essential. After creating a bot, you need to talk to the BOT in Telegram (send a random message) before you can send a message by using API. YYYYYY is Telegram user's ID, you can get it by talking to the bot @userinfobot.
{"message":{"subject":"Server Status Notification","body":{"contentType":"Text","content":"#NEZHA#"},"toRecipients":[{"emailAddress":{"address":"ADDRESS FOR RECEVING EMAILS"}}]}}
Notes for this method: This method requires calling Microsoft Graph V1.0, you need to go to Microsoft Graph and create your own application, give Mail.Send permission and get the Token, or you can go Microsoft Graph Explorer directly to give permission and get the Token, just replace the Token in the Header with the actual Token.
Wechat Work Group bot Example, contributed by @ChowRex
Type: one or more types can be selected, such as in a rule to select more than one type, you need to meet all the selected types at the same time to trigger the notification (see the example later)
cpu、memory、swap、disk
net_in_speed Inbound speed, net_out_speed Outbound speed, net_all_speed Inbound + Outbound speed, transfer_in Inbound Transfer, transfer_out Outbound Transfer, transfer_all Total Transfer
offline Offline monitoring
load1、load5、load15 Load
process_count Number of processes Currently, counting the number of processes takes up too many resources and is not supported at the moment
tcp_conn_count、udp_conn_count Number of connections
duration：Lasting for a few seconds, the notification will only be triggered when the sampling record reaches 30% or more within a few seconds
min/max
Transfer, network speed, and other values of the same type. Unit is byte (1KB=1024B，1MB = 1024*1024B)
Memory, hard disk, CPU. units are usage percentages
Send specific server notifications to specific notification groups
Case: You have four servers, 1, 2, 3, 4, and two different notification groups, A and B 1, 2 The two servers are down for 10 minutes and send a notification to Notification Group A 3, 4 These two servers are down for ten minutes and then send a notification to Notification Group B
First you need to set up two notification groups, A and B, and then add two alarm rules:
transfer_in_cycle Inbound transfer during the cycle
transfer_out_cycle Outbound transfer during the cycle
transfer_all_cycle The sum of inbound and outbound transfer during the cycle
cycle_start Start date of the statistical cycle (can be the start date of your server's billing cycle), the time format is RFC3339, for example, the format in Beijing time zone is 2022-01-11T08:00:00.00+08:00
cycle_interval Interval time cycle (For example, if the cycle is in days and the value is 7, it means that the statistics are counted every 7 days)
min/max, cover, ignore Please refer to the basic rules to configure
Example:
The servers with ID 3 and 4 (defined in the ignore) are counted on the 1st of each month, and a notification is triggered when the monthly outbound transfer reaches 1TB during the cycle.
Always triggered: A notification is triggered each time the status reported by the Agent matches the rules of the notification
Triggered only once: only one notification is triggered when the state changes, such as changing from normal state to abnormal state or abnormal state back to normal state
If you need to perform a task while sending a notification message, you can set those items
Tasks to be triggered in case of failure The task to be executed when the notification status matches the change from "normal" to "failure", the task should be set in advance in the tasks page
Tasks to be triggered after fault recovery The task to be executed when the notification status matches the change from " failure " to " normal ", the task should be set in advance in the tasks page
`,27),l=[a];function p(r,c,i,u,D,y){return s(),e("div",null,l)}const C=o(t,[["render",p]]);export{F as __pageData,C as default};
