Infrastructure and server state control is a responsibility of every administrator. Awareness of state of all devices we use in our everyday work is extremely important to us. This is primarily about the operation of servers, applications, specific services installed on virtual machines, and also the correctness of the applications and production environments. An excellent tool for monitoring this type of thing is Zabbix. How it works?
We have our agents
Zabbix can monitor thousands of data entries from servers, virtual machines, applications, network devices in real time. This allows you to detect problems before the problem is returned to users. On each of the servers that we want to monitor, Zabbix Agent is installed – a special program that allows you to communicate with the Zabbix root server. The agent collects data that is of interest to us, such as current processor usage, RAM, network traffic, database load, performance of individual services (eg http, ssh, ftp). Later on, it sends the collected information to the main server – Zabbix Server, then formulates it clear and easy to read tables or graphs. Data is stored in relational databases (MySQL, Oracle or PostreSQL), and data access is via an intuitive web interface. Since version 3.0, Zabbix supports encrypting of the communication between the server and agents, so we guarantee that all data during the “journey” is safe.
Preview in the command center
Do you associate security guards in some stores who are bored sitting in front of a CCTV monitor? You can implement a similar solution in your company to monitor your hardware and software. It is very easy to generate relevant charts or reports. This gives us continuous control over the performance of all our servers and services, and we can keep an eye on what’s going on.
Zabbix will tell us about the malfunction in many ways. One of the most popular is sending information by e-mail, or you can also send SMS (eg, certain hours). It’s a very comfortable and effective solution to quickly find out about the problem and react to it. Additionally, at Studio Software we have programmed the whole thing so that, in the event of critical problems, the alarm will be triggered. And we mean it literally, because the API (which offers Zabbix) has an Arduino microcontroller connected to a “police cock”. The whole thing is asking Zabbix whether our network is working properly and if anything goes wrong with the server, Arduino turns on the cock installed on the wall. Sirens are out and SysOps are immediately aware of the problem. Fortunately, this happens very seldom, thanks to the corresponding configuration, which is described below 🙂
Important and more important issue
Zabbix can be configured according to our needs and preferences. Although every problem, even the smallest one, is in a sense relevant, they are obviously a subject of a certain classification. With the Zabbix Trigger, the tool will alert us about the problem in an adequate way to his or hers significance. Example? During configuration, we set that less than 20% of free disk space is a average problem. If disk space is less than 10%, then this is a severe problem. Based on the above triggers, the system responds accordingly. If we have agreed that the average problem does not require an immediate response, then we only need a warning mail, which will reach us at a convenient time, so we do not need to worry about it unnecessarily. With crucial news you should be warned immediately, so Zabbix will choose SMS as a contact, or – as we mentioned above – will activate the alarm siren.
Zabbix is a powerful, reliable and highly needed tool. The security of our network infrastructure is extremely important. It is based on our daily work. Let’s stick to the old truth that better prevent than cure.