Zabbix Nagios Non-performance-based development and monitoring it's all nonsense.

From the initial Nagios to the now powerful Zabbix 3.0, I would say that no performance-based development and monitoring are all nonsense?

First of all, I am aware of these two monitoring software:

Zabbix, a lot of enterprises are in use, give a person's feeling is very dazzling, but I personally think that although it is powerful, can draw a cool image, but in you see his "beautiful" face, whether he thought of the original library, the system of "harm", whether have considered some of the indicators really have practical significance.

Zabbix Monitoring server: Through the agent, the relevant data is inserted into the database JSON, and then through the application to the content of the database to be displayed.

Zabbix monitoring database: Through the database user (grant the relevant permissions) directly in the database to establish the relevant query, the direct data is cured locally, and then through the application to show the content.

Nagios, a lot of foreign enterprises in the use of, the most native things, through the full custom script to achieve data transmission to the local and display, do not need to use the database for curing, in my opinion, the purpose of monitoring is to detect potential risks and timely send warnings, and the pure ecological Nagios The core can do this perfectly, and of course Nagios always regrets that it requires the configuration file to be configured purely manually without using the database.

Nagios Monitoring server: Nagios server-side active query, direct display (the most original ecological monitoring)

Nagios Monitoring database: Nagios server actively sends requests to the agent to fetch and transmit the database to the server and show it.

Of course, for really effective and meaningful monitoring is to monitor for certain indicators, of course, such indicators can only be determined by your own system, database understanding, at least for example, the database table space, server disk utilization and other monitoring interval can be set to 6 hours or even 12 hours, etc. Oracle database 10g, 11g version of the memory parameters of the definition of the hit rate and validity.

Development is also the same as the development of the same function, some people can complete a bunch of bugs, and some people can avoid many bugs.

The last word, not performance-based development and monitoring are nonsense.

Zabbix Nagios Non-performance-based development and monitoring it's all nonsense.