Zabbix 4.2 with new equipment "tuned" us for monitoring for comfort and performance

The latest version focuses on high-frequency distributed monitoring and provides new options for monitoring of container, application and cloud resources.

„I am thrilled that Zabbix has been successful in supporting thousands of companies on their path to digital transformation by providing a highly reliable and flexible monitoring solution. The new version sets the bar of IT infrastructure monitoring to a higher level and puts Zabbix in one of the most important components of modern AIOps platforms," said Alexei Vladishev, founder and CEO of Zabbix.

What's new in Zabbix 4.2?

Here is a list of the most important features included in the new version:

Official support for new platforms

In addition to existing official packages and appliances, Zabbix 4.2 now also provides packages for other platforms: RaspberryPi, SUSE Enterprise Linux, Mac OS/X, MSI pre systém Windows, Docker.

Built-in support for data collection from the Prometheus project

Zabbix offers integration with Prometheus exporters using native PromQL language support. In addition, the use of dependent metrics gives us the ability to collect a huge number of Prometheus metrics in a highly efficient way: we retrieve all the data with a single HTTP call and then apply it again to the corresponding dependent metrics. Zabbix can also transform Prometheus data into JSON format, which can be used directly at low level discovery (LLD).

Effective high frequency monitoring

The high monitoring frequency very often overloads the monitoring system. This can be avoided at the input preprocessing level because it allows us to skip repetitive values. We can now collect high-frequency data and immediately detect problems without storing excessive amounts of history data in the Zabbix database, thanks to the "throttling with heathbeat" function, which is actually a deduplication of recording the same measurement values with regular values from a larger time section for creating continuous graphs.

Verification of collected data and error handling

Nobody wants to collect incorrect data. With Zabbix 4.2, we can solve this through built-in input preprocessing rules, which verify whether the data is identical or not with a regular expression, or using JSONPath or XMLPath. It is now possible to obtain a list of error messages from the collcted data. This can be especially useful if we receive errors from external APIs.

Data preprocessing with JavaScript 

In Zabbix 4.2, you can take full advantage of the performance of user-defined scripts in JavaSript. JavaScript support provides absolute freedom to process data. In fact, you can now replace all external scripts with JavaScript. This allows for all kinds of transformation, aggregation, data filtering, arithmetic and logic operations and much more.

Process millions of metrics per second

Prior to version 4.2, all input preprocessing was handled exclusively by the Zabbix server. Zabbix provides greater scalability, as all input preprocessing functions can now be performed via the Zabbix proxy. The combination of input preprocessing performed on proxy servers and the functionality of "throttling with heathbeat" duplicate data allows us tu perform high-frequency monitoring collecting millions of values per second without overloading the Zabbix server.

Simple Low-Level Discovery (LLD)

Zabbix 4.2 supports processing based on JSON input, which allows us to communicate directly with external APIs and use the received data to automatically create guests, metrics and triggers. Combined with JavaScript preprocessing, it opens up fantastic options for templates that can work with a variety of external data sources, such as cloud APIs, application APIs, XML data, JSON, or any other format.