"You must be crazy if you don't use the potential of Open Source." Gerhard Lausser on open source monitoring systems and enterprise solutions

Gerhard Lausser is an expert in the field of nagios-compatible monitoring systems, the author of numerous publications, the developer of widely used plug-ins for monitoring many commercial systems (one of them is check_hpasm, we even analyzed in detail earlier). The interview was performed by an employee of WürthPhoenix, which manufactures monitoring systems in the form of aplliance under the NetEye trademark and published on the company's blog ^* in 2010. Mr. Lausser outlines a look at using monitoring systems to monitor the scale of an enterprise.
_{* This is not an advertisement, NetEye products and WürthPhoenix products are not supplied in the Russian Federation.}


In 2009, you published a book about Nagios, its properties and functional features: what was the main thing in it and what was your experience, considering the plugins you developed?
The purpose of my book was not another rewriting of the official documentation for Nagios. I created it for system administrators who already have enough experience in running and working with Nagios. I wanted to write something like a collection of recipes that will allow them to take the next step. From testing basic functions to monitoring covering the needs of a large company. Today it is not enough to look only at the connectivity of the network infrastructure and the use of disk space. It is no longer enough to check whether your database or web server responds to requests. Business is time critical, so I described how to perform a deep check of the infrastructure. My book focuses on the use of plug-ins for monitoring enterprise-wide systems, such as DBMS, web servers, SAP system, logs. Some time ago, I started developing plugins that do just that, so I use my “knowledge of the problems from the inside” to help users get the most out of their companies.


Let's look from the perspective of an IT manager: what are the benefits of implementing monitoring based on Open Source - free software? What is your professional experience with this?
And here you should not think like an IT manager: implementing monitoring based on OpenSource gives you quality too, but for less money. Except when you have special requirements that only a certain proprietary product can fulfill, you will be crazy enough not to take advantage of free software. If your IT staff is qualified enough, then all your investments are limited only by their working time. If you do not have time, then there are companies that provide relevant services and will gladly help you implement and customize the solution based on free monitoring software, and if you don’t want to monitor the status of your system yourself, the company will do it for you. . You can have a full package of services with support and automatic software updates.
')
You may ask: Where is the benefit? The answer is simple: you don’t have to pay for commercial software licenses. If you look at the monthly price of one monitoring object, it turns out that we are talking about thousands of euros, and sometimes hundreds of thousands. From my personal experience, I can say that wherever we replaced proprietary monitoring systems with solutions based on free software based on Nagios, our clients could not believe what foolishness they had done before and how much money they spent only on software licenses. Free monitoring solutions eliminate most of your problems out of the box, and the rest can be solved by spending a small part of what you spent on licenses.

For many years, free software and large companies represented two incompatible worlds, but now something is changing. What do you think about the current situation and future prospects?
I previously worked in large companies and there I met a fairly large number of free software, but sometimes these were “illegal” installations. Administrators have always known free software and its capabilities, but the management did not know how to handle it. Today, these admins themselves have become managers and most of those who make decisions have grown up on Linux, so they have a pretty good idea of ​​what OpenSource is. It will be difficult for commercial monitoring systems to compete with their free opponents, because even for free software you can get service and support at the highest level.

Taking this into account, some of your plugins were designed to test Oracle DB and SQL servers in complex infrastructures. What is your experience in this area?
The first in the plugin family appeared check_oracle_health. And I wrote it not for fun - a few years ago, I was approached by one of the clients who was looking for a replacement for the Oracle Enterprise Manager software that was used to monitor all of its Oracle database servers. He wanted to solve two problems: to save money, and secondly - to standardize the monitoring landscape. He already worked with Nagios, with which he monitored the state of the network infrastructure and the state of the operating systems, and he did not want to have another separate management system to monitor the DBMS. He asked to implement the functionality of the Oracle Enterprise Manager in Nagios as much as possible. This is how the check_oracle_health plugin came about. Today I receive mail from users from all over the world, and it seems that check_oracle_health is becoming the de facto standard in monitoring Oracle database servers.

This encouraged me to rework the code and write plugins for monitoring MySQL, DB2 and MS SQL servers. They also had great success and many experts use them in their work. Check_mssql_health has become especially popular, as it allowed to build a bridge between Nagios and Microsoft industrial databases.

What do you think about the potential of Nagios in comparison with more “younger” solutions like Icinga or Shinken?
I believe that the development of the kernel Nagios has come to its logical end. At least, I do not expect from him a new functional. The general concept of Nagios has been tested and tested for many years, and everything that it lacked was implemented in additions. Now I see the Nagios core becoming the basis for several commercial monitoring systems, which consist of open source components with optional support, compiled and packaged in packages, like your NetEye or OP5. There is nothing wrong, because customers do not mind spending money on a combination of free software and its support.

On the other hand, there are other, incompatible with nagios, open source monitoring systems and their share of the “market” is growing. If the Nagios ecosystem wants to stay at the top of the list, then long-term innovations should emerge. In order for potential users to receive an explicit signal, a new one must appear not only in additions around the core - changes must occur in the core itself.
That is why Icinga and Shinken entered the game. Icinga focuses more on the package as a whole. A lot of work has been invested in their powerful icinga-web interface. Shinken, on the other hand, is a completely new monitoring system. This is Nagios, rewritten in a modern programming language, the purpose of which is to significantly facilitate further development, and in addition, the application has been redesigned, which makes it easy to create distributed, balanced, and fail-safe installations based on it.

Regardless of the technology used, Shinken is fully compatible with Nagios, which means that you do not need to make changes to the configurations and plug-ins used.

All of them - Nagios, Icinga and Shinken have their advantages, maybe they can coexist, maybe they will inspire each other, or, in the end, only one will remain alive. The open source community will decide what is best.

Many free software companies use direct user interest groups (User Group) to establish direct connections with their customers. What do you think of this strategy?
Being open and talking to people, looking into your eyes is the only way to gain trust for you and your product. Customers are smart, they will not spend money on promises; they compare, draw conclusions and want to feel comfortable. The purchase decision is no longer made on the basis of colorful glossy brochures. The organization or participation of users in the activities of OpenSource, shows the whole community that your company is not a fiend of devil, hungry only for money. Instead, you have the opportunity to familiarize yourself with the problems of users, as well as their ideas, and get an immediate response. Sharing knowledge and stories from their own experience, your users know that you, too, are not all simple, also have problems, but you work hard to solve them. Previously, companies were experts, and users were unknowing beginners. Today, there are highly professional users among your customers who will not treat you well just for knowledge, but since you communicate with them, they start respecting you for making your users' lives easier by doing the work they could would do it yourself, but they have no time for this.

You participated in the Nagios conference in Bolzano and you had the opportunity to familiarize yourself with the NetEye monitoring system. What are your impressions and what do you think about this decision?

First of all, thank you for organizing this event. I liked coming to Bolzano, I am here for the second time, and not only because you have such a beautiful, sunny weather here. As I said, Nagios is increasingly becoming the basis for boxed monitoring systems. NetEye, being one of them, saves the administrator a lot of time associated with installation, compilation, configuration and support. As a NetEye user, you can monitor everything, from the state of the equipment to the level of business processes, and you have reports, inventory, troubleshooting and configuration in one box. And you have a partner who can help you if you need some special features.

In my opinion, this is how monitoring systems should work today.

Mr. Lausser, thank you for the interview.

PS

Plugins for monitoring DBMS can be found here:

check_oracle_health - Oracle SQL Server
check_mssql_health - Microsoft SQL Server
check_mysql_health - MySQL Server
check_db2_health - IBM DB2 SQL Server