Shulou(Shulou.com)06/01 Report--

Official site http://www.zabbix.com/

Zabbix version

LTS:long time support long-term maintenance version

2.2 3.0

Properties:

Data acquisition:

Snmp ssh/telnet agent ipmi jxm (Java Management eXtensions).... so on

Custom sampling mechanism: UserPrameter

Alarm:

Alarm upgrade mechanism:

Execute script: script

Notification: report the failure to the relevant personnel, if the problem is not solved, report to the upgrade supervisor.

Data storage:

Data storage: MySQL/PGSQL # quotes where there are people, where there are people, there are mobile phones

Show:

Real-time drawing: visual data graph (chart)-- > > screen-- > > slide show (slide switching)

Map-- > > create a topology diagram-- > > each node is connected by a line. The node is broken and use the red line and so on.

Support templates:

Automatic network discovery: automatically discover hardware and apply templates to automatically monitor

Distributed monitoring: when the cluster is very large, it needs to be used

ServerProxyagent/ssh/ipmi

API: self-development of development interface

Zabbix: program component

Core components:

Zabbix server: server daemon: responsible for issuing monitoring requests, collecting data and storing them in Zabbix database service master component # C

Zabbix agentd:agent daemon # monitored host

Zabbix web GUI: responsible for retrieving processing data from Zabbix database and displaying key components of the management structure # PHP

Zabbix database: database, carrier for a variety of databases. There is a special script to generate the database structure # the root Zabbix is independent of itself

Zabbix proxy: when the cluster is too large, the agent is responsible for sharing the load. Send data summary to Zabbix server # optional

Zabbix get: a command line tool for manual testing to initiate a data collection request to agent

Zabbix sender: command line tool. Runs on the agent side. Manually send collected data to server

Java gateway used in Zabbix jave gateway:JMX monitoring # do not understand

(Web pager: collect and evaluate the response speed of requested page resources to ensure user experience)

Zabbix monitoring mode:

Passive mode: request by Zabbix server. Passive reply on agent side

Active mode: the agent responds actively to collect information to Zabbix.

*

Zabbix architecture:

Server100: monitored host # there is no relationship between DB and Zabbix here

Server1:Zabbix main program server

Sever2:Zabbix web php

Server3: database server

* * Don't say you don't understand *

*

Common Zabbix terms:

Host (host): each server monitored by Zabbix can be specified by an IP or DNS name

Host group (host groups): usually can be divided into; similar host, divided by function, service, geographic location, structure, etc., divided by business logic, combined (usually the same service, easy to manage)

Monitoring item (item): data related to a specific monitoring indicator. # "key" can be understood as a function item: as a parameter.

Item is defined by key, and every key defines a collection program.

Flip-flop (trigger): an expression used to match whether the item is within a reasonable interval, that is, the gate value; when the value is greater than the gate value, "ok (trigger is triggered)"-- > 'problem'; "problem"-- > >' ok' # conforms to the expression that there is a problem and does not conform to the expression is normal. See the definition of the expression.

Event: an event of concern, such as a trigger state change, a new agent online or offline.

Action (action): refers to the processing action for a specific time # send a notification or restart the service or execute a specified script

Alarm upgrade (escalation):

*

Zabbix logical architecture:

Multiple host form a host group, which is matched by a defined items. When a change triggers a trigger, an event (Events) is generated, and Actions can monitor the occurrence of the event to complete the defined action (Actions) to execute remote commands or send notifications (Email is just one of the manifestations).

If you want to make a quick monitoring configuration for the host, you can use the template (Template), and then apply

Maintenance (maintenance), when in maintenance mode, does not trigger monitoring

Items must correspond to the data collector (Zabbix Poller) pull tool (based on the above three channels)

*

Zabbix server process

When starting the Zabbix service, you need to start so many of these processes. And there is more than one process.

Watchdog: check the processes of Zabbix to prevent them from being kill (watchdog works in the kernel, so the chance of being kill is greatly reduced)

Housekeeper: mainly used to manage the historical data in the database. Expired data is deleted automatically.

Alerter: alarm tool

Escalator: alarm upgrade

Pinger: used to detect whether the host is online

Poller: implement the data collection of item (the number of processes does not necessarily depend on the number of collection items defined)

Monitored by httppoller:web.

Discoverer: automatically discovers hosts

Nodewatcher: monitoring nod

Timer: timer

Db_config_syncer: data synchronization components that complete db configuration synchronization to # proxy db and Zabbix server

The data collected by proxy is not sent directly to Zabbix server but needs to be synchronized.

Db_data_syner: complete db data synchronization

* * configuration requirements *

Generally speaking, the bottleneck of Zabbix is disk I-bank O or network I-bank O.

The hardware configuration is slightly..

DB requirements. Slightly; slightly.

Front-end requirements:

Three major components

Zabbix Server

Zabbix web GUI

Zabbix Database

The above components can be monitored on three different hosts

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.