None IT specialist
doesn’t have the heart to say that the management of computer networks is an "easy job". In recent years, there has been a huge leap forward in the tasks of universal computerization, which adds work for modern system administrators. If earlier small networks were mainly used, today the number of machines and their chains exceeds hundreds, or even thousands of devices.
It is not possible to manually configure them, add or configure the operating system, otherwise the task forces many specialists to be involved in the process. The last decision has its negative sides, because all admins are people and have special thoughts about how best to fulfill the instructions of the management. The human factor inherent in each of us plays a role. And if you entrust the task to one specialist, you need to have a lot of time to spare.
This situation gave rise to the decision to automate this process. Today, specialized software
is actively used, which helps to make OS settings in automatic mode so that the network functions correctly. The task is carried out when there is a remote contact. One of the requirements for this is cross-platform, because different Linux, FreeBSD and Windows operating systems can be used within a single network. In this case, an effective solution would be to use a special network administrator tool that will help manage the OS configuration within a single network.
Network management tools
There are a number of applications that network administrators can use to manage different operating systems. But most often, experts use Chef vs Puppet vs Ansible vs Saltstack
programs. If you start asking them which application is the best in practice, you may come across different opinions on this matter. We propose to figure out together what these programs are, and why admins prefer them.
Chef vs Puppet vs Ansible vs Saltstack: what is it and what's it for?
Each of the programs is designed to manage the OS configuration within one network. Represented by client-server architecture. The administrator, who is located behind the server, is able to send configuration data to the peripheral machines. The client side must be installed on the devices. The system operates automatically, configuring, relying on the settings that the server delivered.
These software developments are used to administer several VPS, manage server fleet, configure their servers from "0", connect additional nodes to the cluster and other important tasks entrusted to the network administrator.
The need to use network automation systems in 2022
Chef vs Puppet vs Ansible vs Saltstack
are created to manage large infrastructures with 10,000 servers on Google, LinkedIn and WikiMedia.
Just imagine that you have several servers and have to do something with them. The algorithm of actions in such a situation provides for logging into each server and performing actions in turn. If a more difficult task arises, for example, installing software
, then you will need to configure it, relying on certain characteristics. The task becomes more difficult, but it will be even harder if you have 10 or even 100 servers at your disposal.
Now you need to log in to each of them in turn - 1, 2, 3... And 100 is still far away. After completing the installation command, we proceed to editing the files with configuration settings, and the process is repeated on each machine. Tiring... Right?!
To avoid these difficulties, you can update the servers at the same time using one of the software products. Let's choose which one is better to use in 2022 – Chef vs Puppet vs Ansible vs Saltstack
. To do this, we will review each configuration management system separately.
In terms of functionality, Chef has similar characteristics to the aforementioned systems, but there are also differences, which are the main reasons for its great popularity among network engineers.
The program is implemented on the client side of Ruby. To manage Chef, you need to have a minimum basic knowledge of this programming language, and install the appropriate software environment on the central server.
The administrator deploys Chef on a workstation, creating a chef server where the main control part of the software is installed. He writes a configuration file - Receipt, to configure the desired direction of the network behavior on all Nodes units. This data goes to the repository, where software files, templates of future configuration characteristics are also sent. The system allows you to write several Receipts in the Cookbook at once. This is a big plus, because there are different versions of the OS and software on the nodes. And there can be more than one storage, which allows you to record different scenarios of network behavior
Network units request current scripts through Chef from the server and are able to accept commands. This is followed by the process of reconfiguring the node parameters.
Scheme of work
It turns out that the Chef system is an excellent solution for efficient and fast network reconfiguration for the tasks that the administrator needs to achieve. The platform is ultimatum, but in order to work with it, you need to gain experience, sharpen skills. But labor costs will be not for nothing. A network engineer will be able to save a lot of time by avoiding annoying mistakes and problems in setting up nodes even on a hundred machines at the same time.
Chef can work in 2 modes:
- offline configuration «chef-solo».
The system is most commonly used for Linux PCs, but has also performed well with the latest versions of Windows
It is one of the most sought-after brands in the CM market, which has been actively developing since 2005. The system is used by such eminent corporations as PayPal, Google, Reddit, Dell, Oracle and Stanford University. This already suggests that Puppet is worth taking a closer look at.
The system is distinguished by the possibility of effective use not only on Linux, Windows, as in the case of the previously considered Chef application, but also on Mac OS X, Unix.
Puppet has Ruby open source. There is professional support and a corporate version of commercial software.
To use Puppet in business, you need to install the server side on the Ruby programming environment. This is where all the system configuration files will be stored. The developer named them "Manifests". During operation, the server will accept requests from client machines. In an automatic process, files with an updated version of the operating system configuration will be sent to them. This will allow all devices to continue to operate efficiently and smoothly on the network.
An important requirement for using Puppet is the installation of the appropriate software on the customer's PC, meaning the client part. These packages are included in the OS. Admin will quickly be able to deploy a computer network to continue performing tasks. If the installation packages are not found, you need to contact the official Puppet website for help. The developer provides guests of the resource with the opportunity to independently download the desired assembly if necessary.
Scheme of work
Another plus of this solution is that one network engineer can configure and manage hundreds or even thousands of computers. If a problem occurs, a response will be triggered. This allows the admin to quickly fix the code error
and fix problems. But here it should be noted that the specialist should be careful in completing the task. If one of the lines is misspelled, it can cause serious problems that affect the entire network.
In fact, if you use the correct manifest of the previous version of the build, this will allow you to quickly restore the chain of machines and bring it into action.
The system has disadvantages. Puppet is another victim of its own fame. Users report bugs that the developer is in no hurry to fix. Some requests are ignored altogether. There is also resentment among brand customers about its insistent hints to purchase a commercial version of the application.
The Ansible brand has surpassed the aforementioned software giant Puppet, gaining 26.5% of the market share in its influential hands, while Puppet has only 12%. It's about the configuration management systems industry.
It is an open source Python product released in 2012. Supported by the AnsibleWorks developer. Python gives the system an advantage, because this language allows applications to run at high speed, it is built into Unix and Linux.
The main advantage of Ansible is easy and fast deployment, it does not even need to use agents to communicate with the master client, the functions are carried out using SSH. If there are configurations that do not support this root profile, the system runs "sudo" as root.
You can run Ansible for simple tasks from the CLI without using configuration files. In this case, we are talking about checking the operation of the service, rebooting and starting updates. If these are more complex goals, Ansible configuration is processed using YAML syntax in the "playbooks" files.
Scheme of work
You can set commands to the system in different programming languages. The most common are generic JSON modules. This is a kind of advantage, because in other applications there is no such wide choice in languages.
The popularity of Ansible is due to the fact that the system developer takes a new approach to performing common configuration tasks. The application is actively used to deploy large data centers in modern companies in the world.
The system developer himself regularly develops the product, increasing the audience of its users. There has been improvement of quantity indicators of supported Ansible devices, integration of better Windows support, ecosystem improvement.
The disadvantages include the fact that the system is still at the stage of developing a graphical interface, and is also relatively young, having no experience of mass testing by users.
The software is open source and written in Python, which is similar to Ansible. Provides a mechanism for remote work. Push-model and SSH protocol are used to execute commands.
It is distinguished by the performance of the assigned tasks at high speed, which outmatches even the abovementioned systems. You can manage configuration, infrastructure, and orchestration. This app is known since 2011.
It’s created with master-slave architecture. The network admin in the application is called Salt Master - he is also the main demon that controls all Salt Minions. He gives one of them or several Minions commands, which they must obey unquestioningly. The received data about changes is delivered to the Master. Communication between the main parties is carried out using ZeroMQ message keys. When connecting to the Master for the first time, the Minion automatically saves the key on the server. There is also a Salt SSH format - "agent-free" management.
Scheme of work
IT automation using SaltStack can be carried out in the following format: one Master - up to 10,000 Minions. The app is as easy to use as it is to customize. The remote execution architecture is uniform. There is support for all kinds of languages by system configuration files.
Another feature of SaltStack is the ability to execute commands in parallel on remote systems, which helps speed up the automation process. The application has a simple interface, works well with BSD Unix, Windows, VMware vSphere.
The main advantages of SaltStack are the reliability of the system, its powerful management structure that can handle 10,000 machines. It is a secure product that uses an encrypted protocol and has a fast and light Fast communication bus that guarantees the basis for a remote task execution engine. The process is carried out automatically, the cloud controller Salt Virt is used for this purpose.
Chef vs Puppet vs Ansible vs Saltstack – what is better?
After reading the overview for each system, you can draw the appropriate conclusions:
- Puppet application has the most experience. Unsurprisingly, its developer was able to organize the largest community of its own and build a solid user base. But other programs are not lagging behind as their manufacturers take new approaches, which allows them to join the ranks of their loyal audience.
- All systems have first-class tools that allow you to effectively control the project life cycle of configuration settings management using the command line.
- All 4 applications integrate well with DevOps systems, cloud computing platforms AWS and Azure.
- Versions of systems with open source code are available to users free of charge. Users of commercial versions of applications receive more guarantees of security, stability and reliability. Puppet offers an Enterprise version for purchase, with prices ranging from $ 112 to $ 199 per year. The calculation is carried out for the maintenance of one network node.
- All 3 systems - Chef, Ansible, Saltstack are quite easy to use. Developers put a lot of effort into making sure users don't have problems. The best in this regard is Ansible. But with Puppet, if you do not follow its way of automation, misunderstandings can arise. The developer has realized his problems and is trying to fix them. It's about combining code and data in source files.
- For storing keys in Puppet and SaltStac versions, the popular YAML format is used, which allows the admin to manage configurations efficiently.
In the main question of our article, what is better to use in 2022 for the network admin Chef vs Puppet vs Ansible vs Saltstack
- it is not possible to give a definite answer. The choice always remains solely with the specialist. All of the abovementioned tools have a lot in common, have the proper functional features, which made them the most popular in the IT field. But in order to finally decide, the administrator should analyze the needs for the tasks assigned to him, highlight for himself the advantages of a particular SCM application, and evaluate a pleasant development design for himself.