zabbix official word description zabbix:
-
Monitor any thing for any IT infrastructure solutions, services, applications and resources
-
Monitor anythingSolutions for any kind of IT infrastructure, services, applications, resources
-
An overview of basic monitoring
-
zabbix not monitoring, but monitoring tools to achieve
-
Zabbix-server is a c / s and b / s structure
-
When the server is installed and php7.1 mounting zabbbix conflict: If this is not already installed php7.1 to install the machine on zabbix
-
-
Monitoring of knowledge
-
Why use monitoring
-
1. continuous real-time monitoring system
-
2. The current state of real-time feedback system
-
3. ensure the reliability of Security Services
-
4. To ensure business continuity and stable operation
-
-
If you go to a new company, how to start?
-
1. Hardware Monitoring - routers, switches, firewalls
-
2. The system monitoring --cpu, memory, disk, network, process, tcp
-
3. Service Monitoring --nginx, php, tomcat, redis, memcache, mysql
-
4.web Monitoring - response time, load time, render time
-
5. Log Monitoring --ELK, (collection, storage, analysis, display) log
-
6. Safety Monitoring --firewalld, WAF (nginx + lua), security Po, Niu Dun cloud, security dog
-
-
-
Stand-alone monitor
-
Stand-alone process to check the load and cpu usage
-
Stand-alone memory View
-
Stand-alone disk View
-
Stand-alone view network
-
-
Zabbix introduced Distributed Monitoring System
-
Use shell scripts to monitor the server
-
-
Installation zabbix
-
(Single) -> LAMP
-
(Architecture) -> LAP + MYSQL
-
-
Server port: 10051
-
Client Port: 10050
Base template
-
Custom monitoring thresholds combat
-
Custom Monitor items
-
unit
-
Value Type
-
Value Mapping
-
-
Defined threshold
-
Single Conditions
-
Multi-conditions
-
-
Custom trigger (operation)
-
Custom Alarm | information content (micro-channel messages) sent messages can use the system comes with macro variables corresponding to modify (the official site has introduced macro variable)
-
Custom graphics, polymerization graphics, slides, network topology, Graphtree
-
Custom template (add to the host)
-
The system comes with a default set thresholds to monitor entry to training (level threshold) modified according to production needs
Service monitoring (monitoring of service status page requires inquiry)
-
nginx
-
PHP-fpm
-
mysql
-
tomcat
-
repeat
web monitor
-
Request Time
-
Response time
-
Pages instead of 200 -> trigger alarm
Automated monitoring:
-
Auto discovery (server side network scanning polling found Agent)
-
** Auto discovery: server -> Scan Polling -> IP addresses -> **
-
Automatic discovery: ip, ftp, ssh, web, pop3, imap, tcp
-
Pham Van ip automatic discovery (in two phases: discovery -> Action)
-
szabbix-web zabixx-agent automatic discovery automatic monitoring network defined in (already defined in the configuration file server zabbix-server address)
-
-
-
Action performed by automatic discovery
-
Send a message
-
Add / Remove Host
-
Enable / disable the host
-
Adding to the host group
-
To remove a host from a group
-
Will host links to templates / unlink from a template
-
Execute remote script commands
-
-
Active registration (agent ends the initiative to tell the server-side request to join)
-
zabbix-server must be turned on automatically register -> Options -> (notification | Join monitoring | Apply template)
-
Agent (ServerActive = 10.0.0.61) -> Start -> Automatic added zabbix-server
zabbix-proxy distributed
-
Zabbix-proxy usage scenarios:
-
Monitoring remote locations across the room to solve
-
Monitoring host more performance can not keep up, delay large
-
Solve network instability
zabbix is how to monitor the implementation of
Probably a process monitoring system runs like this:
agentd needs to be installed on the host to be monitored, which is responsible for collecting the data periodically sent zabbix server side, zabbix server data stored in the database, zabbix web to unfold at the front end and drawing the data. Here agentd collect data divided into active and passive modes:
Initiative: agent requests server initiative to get a list of items to monitor and detect data initiative will be submitted to the server within the monitoring item / proxy
Passive: server agent to request a data monitoring items, agent data is returned.
-
Active mode Passive mode: The default is passive mode zabbix-agent
主动模式与被动模式主要是站在zabbix-agent身份来说-
1. passive mode (zabbix-server polling detection zabbix-agent)
-
2. Active Mode (zabbix-agent actively reports to zabbix-server) Excellent
zabbix主动模式与被动模式选择-
1. When (Queue) queue delay monitor a large number of items
-
2. When monitoring host more than 300 +, it is recommended to use active mode
[Active] Monitoring the communication process is as follows:
First zabbix configured to request acquisition ServerActive IP active items, data acquisition and submit active tiems value server or proxy. Many people will ask questions: zabbix long time get active items? It RefreshActiveChecks the frequency profile is, if the acquisition has failed, it will try again after 60 seconds. Divided into two parts:
Get ACTIVE ITEMS list
Agent TCP connection is opened (active detection becomes open Agent)
Agent detection request list items
Server returns a list of items
Processing Agent response
Close the TCP connection
Agent start collecting data
Active submit data detection process is as follows:
Agent establish a TCP connection
Agent to submit a list of data items collected
Server processes the data and returns the response status
Close the TCP connection
[] Passive monitoring communication process is as follows:
Server opens a TCP connection
Server sends a request agent.ping \ n
Agent receives the request and response
Server processes data received
Close the TCP connection
1, when the new monitoring project, chose zabbix agent or zabbix endpoint Agent (active), the former is passive mode, which is the active mode.
2, set agentd profile StartAgents parameters, if 0, indicating that prohibit passive mode, otherwise open. General recommendations Do not set to 0, because a lot of time monitoring project, you can use part of the initiative, part of the passive mode.
zabbix custom discovery is how to do
1, you first need to create a rule to automatically discover them in the template, this place needs only a name and a key.
2, the middle value of the macro you want to add filters used you need.
3, and then to create a prototype monitoring items, but also a name and a key.
4, and then need to write such a collection of keys.
Automatic discovery actually need to go to get the value you want to monitor, then this value as a new parameter passed to another item of data collection to go inside.
zabbix is how micro-channel alarm ---- businesses now use relatively more
1, first of all, need to have a micro letter Enterprise. (A real-name authentication [micro signal] can be used in a [phone number] a log of [mailbox number]
2, download and configure the micro-channel public platform private interface.
3, the configuration Zabbix alarm (warning increased media type, media add user alarm, adding the alarm action).
zabbix how to open a custom monitor
1, write a script to get some status information to be monitored services.
2, add a custom "UserParameter" in zabbix client configuration file zabbix_agentd.conf, the aim is to facilitate zabbix call the script we wrote above, to obtain information to be monitored services.
3, in zabbix server can test whether zabbix_get zabbix to acquire the data collected by the client defined parameters of the second step.
4, the new service at the end of zabbix web interface templates, while the first step of what the script can get information on what to add monitored items, "key" set to the value "UserParameter" configuration of the front.
5, the graph display data, and graphic directly to the new monitored item selection step to generating dynamic charts.
zabbix monitor how much client clients is how batch installation
Taiwan company based on the actual number of answers.
1, use the command to generate the key.
2, the host sends the public key to all client installation zabbix.
3, software installation ansible, (modify the configuration file, add the client into the group zabbix).
4, create a script zabbix client installation.
5, the implementation of the script.
6, verification.
Practical experience summary:
-
1. First View the document there is no corresponding xml scripts and templates
-
2. Import template on the server, see the corresponding monitoring item name
-
3. Does the test script can value, and store placed in / etc / zabbix / scripts directory, be sure to add execute permissions
-
4. Write xx.conf file, which is stored in the main monitor how to define items
-
The final restart zabbix-agent
-
6. Use server zabbix-get acquired data monitoring items corresponding zabbix-agent
-
example:
-
The company is not enabled swap (swap is not recommended to enable the company server, because the swap disk is to use an analog memory, cpu consumption performance, it is recommended to close swap. Increase the memory), along with increasing customer traffic, leading to the zabbix service process forced OOM, Zabbix server process is kill, there are two ways to solve, if the company in order to increase memory performance sake, if limited company owned add swap, if it is recommended to use efficient services to add memory,
-