Find out what machine room classification levels are available! Make a comparison table!

Comparison of data centre levels

The table below provides a high-level overview of the four data centre levels and shows what the different models offer customers.

Operating time guarantee 99,671% 99,741% 99,982% 99,995%
Annual shutdown <28,8 hours <22 hours <1,6 hours <26,3 minutes
Component redundancy None of the above Partial power and cooling redundancy (partial N+1) Total N+1 Fault tolerant (2N or 2N+1)
Simultaneous maintenance No No In part Yes
Price $ $$ $$$ $$$$
Field clearance No No No Yes
Staff None of the above 1 shift 1+ shift 24/7/365
Typical customer Small firms and start-ups with simple requirements SMEs Growing and large businesses Government bodies and large companies
The main reason why companies choose this level The cheapest data centre level Good cost-performance ratio Fine line between high performance and affordability A fault-tolerant facility, ideal for continuous high traffic or processing needs

What is the difference between a Tier 3 and a Tier 4 machine room?

  • Level 1: Data Centre with a single power and cooling path, without spare parts. The expected operating time of this level is 99.671% per year.
  • Level 2: Data Centre a single path for power and cooling, and some redundant and spare components. This level offers an expected uptime of 99.741% per year.
  • Level 3: Data Centre multiple power and cooling paths, as well as redundant systems that allow staff to work on the installation without taking it offline. The expected lifetime of this level is 99.982% per year.
  • Level 4: Fully fault-tolerant data centre with redundancy of all components. The expected lifetime of this level is 99.995% per year.

Find Tier 3 and Tier 4 machine rooms in Budapest!

  • TIER III: Ivitech
  • TIER IV: Not currently in Budapest, in principle the needs of the region do not support the need. Arubacloud has one in Italy.

Can water be used in the engine room? Justify!

It can't be a default case, because electrical devices and water fought a long time ago and have been ancient enemies ever since. (Water can damage equipment, but new technology is supposed to send water in by atomization and not cause damage. At least that's what they claim.

What extinguishing systems/devices can be used in a machine room?

With powder and gas (halon), and in principle with the atomized water solution.

What is the difference between a "rack mounted" and a "blade" server?

Blade Server vs Rack Server: main differences
Both blade and rack servers are used in data centres or large enterprise environments. Because blade servers have limited computing components, the smaller rack cabinet space can accommodate multiple servers. The rack servers are very wide and short, so a single server can fit in a single section of the rack cabinet. In contrast, a blade server can fit 20 blades in a small area of the rack cabinet.

Blade servers are used for high-performance processing and, unlike rack servers, are hot-swappable. Hot-swappable means that blade servers in a rack can be removed and replaced without shutting down the entire rack. This reduces downtime if an administrator needs to replace a blade server or remove it from the cluster for maintenance.

Rack servers are complete machines, each of which is necessary for individual operation. Each server needs its own cable and peripherals, unlike a blade server, which can be powered by a single cable to the rack. Administrators can install a single rack server to support services and add servers as needed. Rack servers benefit small organizations that are growing into mid-sized businesses and require affordable upgrades from a standard tower.

The cost of humidity control and cooling is higher for rack and blade servers. The more servers you add to the rack cabinet, the more heat is generated, so additional cooling may be required. The upfront cost of blade servers is higher depending on the number of servers initially installed. You can save money by ordering a server and upgrading slowly, but overall, more blade servers cost more than rack servers. A blade chassis can cost up to $8,000, while a rack chassis can cost only a few hundred dollars.

What is the difference between a desktop and a server?

Workstations basically lack redundant deployments, i.e. 1 CPU, 1 power supply, multiple disks but they are not RAIDed. Neither is the motherboard thermal design designed for increased stress. Also, the fault tolerance of some of the main components is lower. Accordingly, the failure of a component will result in an immediate outage.

Workstations at the hardware level do not have the depth of sensor-based monitoring systems that can (if not always) predict failures in the near future.

Also, workstations do not include a management module (HPE: iLO; DELL: iDRAC; IBM: IMM) to manage them remotely (regardless of the operating system). That is, the machine is only remotely accessible when the operating system is running, with the remote access part of its functionality running. If the machine cannot be accessed via OS, it can only be brought back online physically from the console.

In addition, the enclosure is smaller in size and material and does not provide as much physical protection as expected for a server design.

Let's summarise the advantages and disadvantages of using workstations as servers.


  • A very cheap purchase price per unit of CPU power.

  • It takes up less space than a tower server.


  • There is no remote management facility.

  • Monitoring is not available remotely in a way that is separate from the operating system.

  • Low mechanical protection due to the housing design.

  • No redundant power supply.

  • No (or rarely) integrated RAID controller on the motherboard.

  • Implementation is not planned for 5 years of 7×24 hours.

  • HotPlug solutions are not supported.

  • Low manufacturer support (especially for non-brand machines).

Educational material on the HPE blade technology demonstrated the blade system architecture. Look for the DELL equivalent!

DELL Blade servers


Create hardware configuration according to the following requirements on HPE, DELL and Lenovo bases (2 Intel CPUs with minimum 2 GHz and 10 core; 256 GByte RAM; 1 TByte SSD disk in RAID1; 16 Gbps HBA QLogic; 2x10Gbps NIC)

Go for it! I'll skip this one.

Create hardware configuration according to the following requirements on HPE, DELL and Lenovo bases (2 Intel CPUs with minimum 2 GHz and 10 core; 64 GByte RAM; 240 GByte SSD disk in RAID1, 8 HDDs with 2 TByte in RAID6; 2x10Gbps NIC)

Go for it! I'll skip this one.

Is it possible to install, use and operate MS Windows operating system without a graphical user interface?

Yes, if it is a core version.

Is it possible to install, use and operate a Linux operating system without a graphical user interface?

Yes, it was CentOS 7 minimal, Ubuntu Server- Lightweight Minimal, Fedora server, today maybe only Debian Stable Server has a dedicated command-line UI. But nowadays it's common for these to have a graphical user interface, they just don't use it.

Can I run an application running on MS Windows Server on a Linux operating system?

Yes. This can be done with the Wine package. Version 7.0 was released this year. Also, some distros include it pre-installed, e.g. Zorin OS, SteamOS.

Is it necessary to install a driver under MS Windows OSes?

When you install Windows, you should always install the necessary drivers for your components. Motherboard chipset drivers, graphics drivers, wifi drivers and more. Fortunately, Windows makes life easy for people by delivering most of these drivers to users in the form of Windows Updates. And when you connect a new device to your computer, it gets the necessary drivers from the internet. So if a piece of hardware doesn't work on Windows, in most cases there is a way to make it work. Most non-functioning hardware is just a driver away from working.

Is it necessary to install a driver under Linux OSes?

On Linux, drivers work a bit differently. Most of the drivers for your computer's hardware are open source and are built into the Linux system from the start. Drivers are usually an integral part of the kernel, although some graphics drivers are part of Xorg (the graphics system) and printer drivers come with CUPS (the printer system). This means that most of the available drivers are already on the machine. These are often developed as a hobby by the Linux community but often fortunately they are also developed by the vendors and are made available to Linux kernels and other projects. So in summary, with Linux, most of the drivers you need are already on the system, you don't have to hunt for them. The Linux system will automatically detect the hardware and use the drivers required for it. However, it is still up to the user to obtain proprietary drivers, even on Linux. The vast majority of these are produced by manufacturers and are closed source, so will not be available by default on Linux distros. Examples include AMD, Nvidia and other graphics drivers that provide a better gaming experience on Linux. There are, of course, open source graphics drivers, but they do not offer the same features as AMD or Nvidia drivers. The method of installing proprietary drivers depends on the distro. On Ubuntu and Ubuntu-based distros there is an "Additional drivers" tool. This can be brought up in the search engine with a few clicks and 1 search. It checks which proprietary drivers you can download for your current hardware. On Linux Mint, for example, the ''Driver Manager'' does the same job. The Fedora distro, on the other hand, is already against such drivers and does not make it easy to install them. So each distro has its own way to install drivers. Have a successful installation!

Under MS Windows, which application can be used to manage the registry?

Yes, with the built-in application called Registry Editor.

Add the Resource Monitor service to the server

I don't fully understand this, because it's a basic built-in service that runs by default. So how do we add it?

You can start it in the following ways:

  • Start menu > Administrative tools > Resource Monitor
  • Windows key + R > type resmon > Run button
  • Server Manager > Tools > Resource Monitor

What is an NTP server? Why is it important to set it up?

NTP (Network Time Protocol) is a standard that ensures that the clocks of machines on a network are synchronised. It is now one of the official Internet protocols and is accurate to a fraction of a second. NTP is very important because accurate time is essential for a network to work: updates, security processes and much more all work on the basis of synchronised time. NTP denotes both the protocol and the client/server application running on the machines. The client requests the exact time from the server at regular intervals and sets its own clock accordingly.

Does NTP communicate on TCP or UDP ports?

NTP uses the UDP user datagram protocol to transmit requests and responses. NTP supports two different modes: NTP Unicast, NTP Multicast.

Why is it important to use Windows Update? How is it used in a large enterprise environment and what are the critical points for updating systems?

Benefits of Windows Update
The main benefit of Windows Update is that it automatically keeps your server up to date with security and bug fixes with little effort. It protects your server from malware and fixes known Windows errors and other problems. It's also an easy way for Microsoft to issue mini-updates when it wants to add new operating systems and software services. It also has the added benefit of checking for updated drivers that are made specifically for your hardware, without having to sort through a long list of devices that don't affect you. Windows Update offers only the updates your server needs, without guesswork.
The challenges of Windows Update
Although Windows Update makes it easier to update server software, it can also cause problems. If updates are not tested properly, they can cause problems on the server due to update errors. Unfortunately, this happens quite often. This is one of the reasons why most administrators download updates automatically, but Windows Update allows them to decide when to install them. This gives the administrator time to check if other administrators have had problems with the update.
In addition, changes made by Windows Update can sometimes cause problems with third-party software. The frequency of updates can also cause problems. As you test and install one update, another one is waiting. It's up to administrators to decide when and which updates are critical and what can wait.

Where can we change the primary and secondary DNS servers?

Now that's an interesting question, because either Settings > Network & Internet > Change adapter options > Ethernet0 > Properties > IPv4 and here you have Preffered DNS server and Alternate DNS server

We got that and it was said that there would be DNS server management later, but that's not the question.

If you look at the question, you need to add the DNS feature in Server Management and then DNS in Tools and configure it there. But this is not included in the current material to my knowledge.

Where can we register our computer as a domain? What authorisation do I need?

Settings > System > About > Rename this PC (advanced) (on the right) > Change settings > Domain (radio button). I think you need an administrator to do this.

Disable Remote Desktop on the firewall

Control Panel\System and Security\Windows Defender Firewall\Allowed apps and uncheck Remote Desktop.

What are the manufacturer's hardware specifications (minimum requirements and compatibility) for CentOS 8?

Minimum system requirements for CentOS 8:

  • 2 GB RAM
  • 2 GHz or higher processor
  • 20 GB hard disk
  • 64-bit x86 system


Create a bootable CentOS 8.2.2004 operating system installation stick!

You can find out how here.

What's the difference between installing with "yum" and "rpm" parameters?

What is RPM

The RPM is a command-line package manager developed by Red Hat in 1995. The package manager was designed to run on Red Hat-based systems. Today, RPM is a core component of many Linux distributions, including CentOS, Fedora, Oracle Linux, openSUSE, Mageia, etc.

RPM Package Manager allows users to query, check, install, update and remove packages. The main disadvantage is that it does not resolve package dependencies or automatic package updates.

What is YUM

A YUM ( Yellow Dog pdater , M odified ) is an open source Linux package management application that uses the RPM package manager. This front-end RPM tool allows users to browse official and third-party repositories and install, update or remove packages from the system.

YUM is the /etc/yum.repos.d/*.repo works with online repositories listed in the file . In addition, the tool allows users to add their own *.repo their files.

The advantages of YUM over RPM are automatic updates, easy package management and dependency management.

It installs both RPM and YUM packages, keeps the information in a database, and provides basic command line functionality. However, there are many differences between the two parcel handlers.

The table below gives an overview of the Main differences between RPM and YUM :

Parameter RPM (Red Hat Package Manager) YUM (Yellow Dog Updater, modified)
Retrieved from It was introduced by Red Hat in 1997. Updated from YUP to YUM in 2003.
Definition Low-level packet manager with basic functions. Top-level front-end package manager with advanced features.
Addictions It does not dissolve addictions. Automatically resolve and install package dependencies.
Package installation Allows multiple package versions to be installed.
However, only one package can be installed at a time.
Does not allow multiple package versions to be installed. Only supports packages available in the repository and shows packages already installed.
On the other hand, YUM can install several packages at the same time.
Updates RPM does not support automatic updates. YUM allows automatic updates to the latest available version.
Repository support RPM does not use an online repository to install packages. Instead, the exact local .rpm package path is required to complete the installation. YUM relies on an online repository to install packages. The utility only requires the package name.
Autonomy RPM is autonomous and uses its own database to store information about packages on the system. YUM is a front-end utility that uses the RPM package manager for package management. The utility also uses the RPM database in the background.
Easy to use The management and administration of RPM packages can become complicated at times. This is the easiest way to manage RPM packages.
Scroll back to RPM does not support rollback of changes. YUM allows you to undo changes.