1, zabbix monitoring
1, zabbix monitoring architecture
zabbix monitoring architecture in the actual monitoring architecture, zabbix according to the network environment, monitor and scale architecture: server-client, server-proxy -client.
1, without any proxy between the simplest architecture server-client architecture also zabbix, the monitoring machine and the monitoring machine, exchange data directly between zabbix server and zabbix agentd. For Network is relatively simple, less equipment monitoring environment.
2, server-proxy-client architecture in which the proxy is a server, a bridge of communication between the client, proxy itself is not the front, but it alone does not store data, but will agentd sent to the temporary storage of data, and then submitted to the server. Generally applicable to medium-sized network architecture across the room, across the network monitoring
master-node-client
2, Zabbix advantage
开源,无软件成本投入
Server 对设备性能要求低
支持设备多,自带多种监控模板
支持分布式集中管理,有自动发现功能,可以实现自动化监控
当监控的 item 比较多服务器队列比较大时可以采用被动状态,被监控客户端主动 从server 端去下载需要监控的item 然后取数据上传到 server 端。 这种方式对服务器的负载比较小。
Api 的支持,方便与其他系统结合
3, Zabbix shortcomings
Data to be installed on the monitored host agent, all the data are present in the database, resulting in a large, major bottleneck in data
Library. 4, Zabbix monitoring system monitoring objects
IPMI: Intelligent Platform Management Interface (Intelligent Platform Management Interface) IPMI can across different operating systems, firmware and hardware platforms, intelligent monitoring, control and automatic return of a large number of operational status of the server, the server system to reduce costs.
SNMP: Network Management Protocol ( the SNMP ) is specifically designed for network management in the IP network nodes ( servers , workstations , routers , switches a standard protocol, etc.), which is an application layer protocol.
5, monitoring the difference
1.nagios graphics are not particularly good, you can also install the graphics plug-ins, but not how good-looking
2.nagios In general, if desired graphics can be used in conjunction with cacti and
monitoring 3.cacti a poll monitoring, low efficiency, graphics relative nagios looks better
4.zabbix and because it is complicated by nagios monitor, cpu requirements for higher
5.zabbix in performance and functionality are a lot stronger
6.zabbix quite beautiful graphics
7. supports a variety of monitoring methods zabbix -agent snmp etc.
8. support for distributed monitoring, can monitor the agent is very much
9.zabbix graphical web configuration interface, configuration simple
10.zabbix support automatic discovery
6, Zabbix monitoring methods
Passive mode
Active mode
1, Server
Zabbix server agent program is reporting system availability, and integrity of the core components of the system statistics, all the configuration information, the memory core operation data and statistics.
2, database storage
All data collected Zabbix configuration information and are stored in a database.
3, Web interface
In order from anywhere and on any platform can easily access Zabbix, we Zabbix Web-based interface. The interface is part of Zabbix Server, usually with Zabbix Server running on the same physical machine.
4, Proxy proxy server
Zabbix proxy can collect performance and availability data for Zabbix Server.
Zabbix Proxy is a monitoring device may be controlled by monitoring data collected from one or more, and transmits the information to the Zabbix sever, basically work on behalf of the server. All data collected will be cached locally and then transferred to Zabbix sever proxy belongs.
Proxy deployment is optional, but can be very beneficial to spread the load of a single Zabbix sever. If only the proxy data collection process on the sever will reduce CPU consumption and disk I / O load.
Zabbix proxy is done in remote areas, offices, ideal for centralized monitoring of the network administrator does not have local solutions.
Zabbix proxy need to use a separate database.
5, Agent monitoring agent
Zabbix agents monitoring agent monitors deployed on the target that actively monitor local resources and applications, and the data collected is reported to Zabbix Server.
6, traffic
monitoring, in order to create a monitoring item (item) for collecting data, you must first create a host (host). Alarm aspects, create a trigger (trigger) in the monitoring item to trigger an alarm action (action) by the trigger (trigger). So, if you want to receive Server XCPU load is too high alarm, you must: 1. Create a host for the Server and associated with a monitor for CPU monitoring item (Item). 2. Create a trigger Trigger, set when the CPU load is too high will trigger 3. Trigger is triggered, sends an alarm message.
7, the meanings of terms used Zabbix
1, the host (host)
- You want to monitor a network equipment, represented by IP or domain name
2, the host group (host group)
- Host logical group; and templates that contain host. And there is no direct correlation between a host and host group in the template. Usually when permission to use the host to the host group assign different user groups.
3, monitored item (item)
- Specific data you want to receive the host, a measure of data.
4, the trigger (Trigger)
- A logical expression is used to define thresholds and issue "Evaluation" monitored item received data
when the received data is above a threshold, triggers the "OK" becomes "Problem" state. When the received data is below a threshold, the flip-flop to retain / return an "OK" state.
5, an operation (action)
- A predefined action to respond to events.
An action by the operator (e.g., notification) and a condition (when operation is occurring) Composition
6, medium (media)
- Means to send an alarm notification; way alarm notification
7, remote command (remote command)
- A case where a pre-defined, to meet certain conditions, the command may be automatically executed on the host to be monitored
8, the template (template)
- A group that can be applied to one or entity (monitored items, triggers, graphics, applications, Web scenes, etc.) on multiple hosts a collection of
templates mission is to accelerate the implementation of the monitoring mission of the host; you can also make the task of monitoring batch modification easier. The patterns are directly related to each individual host.
9, web scene (web scenario)
- Using one or more HTTP requests to check the availability of the site
10, the front end (frontend)
- Zabbix provides a web interface
11、Zabbix server
- Zabbix software to monitor core program, the main function is to interact with the Zabbix server and Agents, trigger computing, send an alarm notification; and centralized data preservation
12、Zabbix agent
Zabbix agent deployed on the target monitoring, active monitoring of local resources and applications (hardware, memory, processor statistics etc.).
Zabbix agent local operation information collected and reported data to the Zabbix server for further processing. Once an exception (such as hard disk space is full or has a service process crashes), Zabbix server will automatically alert administrators to specify an exception on the machine.
