What is OpenStack?

OpenStack is a col­lec­tion of programs that are used to create a cloud en­vi­ron­ment. The term “cloud computing” is used to refer to the de­cen­tral­ized calling up of programs and data. In a cloud, data is saved in several, sometimes phys­i­cal­ly distant locations. This requires an in­fra­struc­ture that allows for de­cen­tral­ized access, in order to guarantee maximum security. Data which may be of a sensitive nature should not only be made highly available, but it should also be protected from unau­tho­rized access and data loss.

Cloud solutions offer the un­beat­able benefit of also func­tion­ing on connected devices with less powerful hardware com­po­nents. The flexibly scalable computing power is in the cloud itself. All that is needed is a fast and stable internet con­nec­tion, which makes setting up a complex network less expensive. There are no high purchase prices at all for the ac­qui­si­tion of hardware. The only costs incurred for working with a cloud solution are normally those related to the actual resources being used. It is also possible to operate a cloud en­vi­ron­ment on rented virtual servers, which also makes it in­ter­est­ing for small busi­ness­es and start-ups that cannot afford large IT expenses.

The OpenStack software pack includes all of the necessary com­po­nents for creating a cloud en­vi­ron­ment. Despite being a rel­a­tive­ly young project, it is still well-developed and secure. OpenStack was created in 2010 by the U.S. space agency NASA and the company Rackspace. NASA wanted software for its project-related work that would make it possible to create flexible and temporary computer networks. Since then, notable companies such as AT&T, RedHat, Canonical (the developer of Ubuntu), Intel, IBM, and Huawei have con­tributed to its de­vel­op­ment.

The software only runs on Linux operating systems, and was published with an Apache License. As such, the source code is openly available to anyone and the program can be used for free without any lim­i­ta­tions. The 20th version of OpenStack was released in October 2019 under the name “Train.”

The OpenStack cloud is made up of several com­po­nents, of which the most important are:

• Nova (compute service)
• Keystone (identity service)
• Glance (image service)
• Neutron (net­work­ing)
• Cinder (block storage)
• Swift (object storage)
• Horizon (dashboard)

Nova is the computing component of the OpenStack cloud and the first one with which OpenStack was launched in 2010. To some extent, it is its backbone, and it is re­spon­si­ble for the most important task: managing the virtual computer. Nova has been developed to the point that it can manage entire groups of virtual computers. In­di­vid­ual computers are synap­ti­cal­ly connected to one another, and the number of nodes varies. To create virtual machines, Nova primarily uses the license-free KVM hy­per­vi­sors that are in­te­grat­ed into the Linux kernel, as well as XEN hy­per­vi­sors that were created by the Uni­ver­si­ty of Cambridge.

Keystone is re­spon­si­ble for au­then­ti­ca­tion and assigning per­mis­sions to users (identity). In cloud computing, not only in­di­vid­ual PCs, but also entire networks are mapped. As such, user au­then­ti­ca­tion and isolating par­ti­tions are very important. Keystone assigns in­di­vid­ual access in­for­ma­tion to each of the cloud’s users (called “roles”) with very explicit rights.

Glance provides image services, meaning images of the data carriers of virtual machines. Glance can also secure and retrieve images. This makes it possible to create a sort of template library for required systems, which can then be generated in the network as often as desired. In addition, Glance ensures avail­abil­i­ty since the required machines can be re­con­struct­ed at any given time.

Neutron (formerly Quantum) provides the virtual network in­fra­struc­ture of OpenStack. It can be used to divide sub-networks, manage IP addresses, and create virtual networks (VLANs) virtual networks (VLANs). Neutron also supports VPN (Virtual Private Net­work­ing). It enables data exchange among the com­po­nents of OpenStack, for example, between in­di­vid­ual virtual machines. Neutron also provides a firewall for the network.

Cinder is re­spon­si­ble for the provision of permanent memory in the form of block storage, like hard drives. Cinder does so through vir­tu­al­iza­tion. As such, the volume can be modified ac­cord­ing­ly (scal­a­bil­i­ty). Cinder block memory functions like a physical hard drive on a computer. Saving data is simple since Cinder makes a user’s hard drives available to them through cen­tral­ized APIs, and it also has a snapshot function.

Swift is the object store. It can integrate dis­trib­uted memory in which seemingly randomly dis­trib­uted data objects can be used on connected drives. Re­dun­dan­cy will be created if needed since the objects can be phys­i­cal­ly stored multiple times. In addition, storage created by Swift can also be used by Cinder or Glance. It is also possible to use dis­trib­uted object storage that has been generated by Ceph or GlusterFS as a sub­struc­ture.

Horizon acts as the dashboard. In other words, Horizon is the graphic user interface through which the com­po­nents in the OpenStack network can be managed. User man­age­ment can also be managed in the process. The design and func­tion­al­i­ty of Horizon are cus­tomiz­able.

The OpenStack cloud is the software-based replica of an in­cred­i­bly complex IT in­fra­struc­ture. The modules that must normally exist phys­i­cal­ly are each virtually generated by one of the afore­men­tioned com­po­nents. Required memory is created by Swift, and then op­tion­al­ly made available as block storage through Cinder. The func­tion­al­i­ty of an in­di­vid­ual server and computer within the network is generated and managed by Nova. The pool of available virtual machines is provided by Glance.

Neutron handles net­work­ing, as well as pro­tec­tion through a firewall and mon­i­tor­ing. Keystone manages the au­then­ti­ca­tion and au­tho­riza­tion of users. Once the system has been built, it can be com­fort­ably ad­min­is­tered through Horizon. If more storage space or per­for­mance is needed for a short period, both of these factors can be adjusted cor­re­spond­ing­ly.

The OpenStack cloud offers the following ad­van­tages:

• Few pre­req­ui­sites are needed to use it: Cloud services can be used on any device with internet access
• High data security and re­li­a­bil­i­ty
• Scalable storage volume and per­for­mance
• Cloud services enable col­lab­o­ra­tion
• Access is not limited by location
• High dis­tri­b­u­tion and ac­cep­tance
• Uniform standards

Un­for­tu­nate­ly, there are also certain dis­ad­van­tages to using OpenStack:

• Highly-dynamic range of functions: OpenStack is an open and very dynamic cloud-computing solution. New functions are added regularly, but some other functions can also be removed.
• OpenStack does not offer any sort of organized support. A lot of doc­u­men­ta­tion and helpful articles can be found online, but it can take quite a bit of time and effort to find the right ones. Unlike com­mer­cial solutions, there is no dedicated customer support.
• KVM and XEN are primarily supported as hy­per­vi­sors, which can lead to problems when in­te­grat­ing other vir­tu­al­iza­tion solutions such as VMWare. For example, users report network problems when using VMWare along with OpenStack.

Due to the absence of a license, using OpenStack is extremely in­ex­pen­sive. Once the system has been set up, it can be easily managed. The use of virtual servers that include hosting services is already possible through a small cloud solution that costs $10 to $20 per month. That means OpenStack can be used as a platform for private cloud ap­pli­ca­tions, for example, a bit like a Wiki. It can even be used to implement sideline projects or even start-up ideas, such as online shops.

OpenStack is an ideal option for SMBs to get started in the cloud-computing world, as well as for the transfer of data and ap­pli­ca­tions. For example, it also offers the pos­si­bil­i­ty of creating a public cloud. Rights man­age­ment within the OpenStack cloud, in par­tic­u­lar, offers a wide range of pos­si­bil­i­ties for optimally managing the col­lab­o­ra­tion of various user profiles.

Large companies can transfer parts of their network in­fra­struc­ture, such as par­tic­u­lar­ly sensitive data or programs, and also use OpenStack within the framework of a multi-cloud approach or use it as a component of a hybrid Cloud. The failure of ap­pli­ca­tion servers is lower as a result since servers can be repli­cat­ed as often as desired. Data is made redundant in any case in the system, so that the loss of data through hardware damage is virtually im­pos­si­ble.

Anyone who would prefer not to use OpenStack can replicate the required services using virtual servers or online storage, for example. Using solutions such as Owncloud or Nextcloud is also possible.

Other similar solutions are:

• Open­Neb­u­la (free)
• VMWare (fee-based)
• Microsoft AzureStack (fee-based)
• Google Anthos (fee-based)

In addition, more and more companies are using the con­ve­nient option of using a service provider to set up and manage a cloud (managed cloud).

No matter what you decide, there are several good arguments for getting into cloud computing. When it comes to ef­fi­cien­cy and flex­i­bil­i­ty, cloud computing is ab­solute­ly un­beat­able. There is also the en­vi­ron­men­tal factor to cloud computing that makes it favorable since the purchase of un­nec­es­sary hardware can be avoided or done without al­to­geth­er.

But the most important factor is def­i­nite­ly the security of your data. It takes a lot of trust to place your own data or that of your clients into the hands of external service providers. As such, the effort of creating your own cloud with an es­tab­lished product like OpenStack can pay off very quickly.