openstack case study

Exploring OpenStack cloud case studies

Opensource.com

During the course of the last twelve months, the OpenStack community has advanced as more users of the leading open source cloud technology have been reporting their progress—with the help of their partners—towards making a meaningful impact on their business goals and objectives.

We've also learned how these progressive technology users are pioneering changes in their own organizations—enabling them to become more competitive in the global networked economy.

The OpenStack Summit 2014 opened with a keynote on the first day of the conference featuring lessons learned from industry leading organizations that have already deployed these open cloud technologies. Each case study had multiple instances of OpenStack that have been applied for a variety of commercial application scenarios.

Jonathan Bryce, the Executive Director of the Foundation, introduced the growing community of  OpenStack Superusers . Glenn Ferguson, Head of Private of Cloud Enablement at Wells Fargo Bank, and Chris Launey, Director of Cloud Services & Architecture at the Walt Disney Company, described their IT environment and how they’ve adopted OpenStack as a solution to their company’s varied cloud infrastructure requirements.

Addressing the needs of enterprise workloads is only part of the evolving adoption story told at this year’s Summit. Several broadband service providers are also early adopters of OpenStack.

How network operators apply OpenStack platforms

During a lunch session on the same day, Fernando “Fred” Oliveira, cloud architect at Verizon, described his experience building OpenStack clouds. He explained how the company’s network is growing faster than they can monetize the infrastructure, and so they’re looking at open source technology as a way to proactively control costs—but their motivation goes beyond bottom-line savings.

Verizon will use OpenStack as a key enabler of its go-to-market business strategy—essentially compressing the time it typically takes them to reach significant new service revenue attainment.

Moreover, Verizon looked to this rapidly evolving software technology as a way to become more agile—from the initial service launch to the ongoing operational requirements. They have been a pioneer of  Network Functions Virtualization  (NFV), and OpenStack became a natural choice to use within their numerous proof-of-concept tests and ongoing pilot projects over the last year.

Exploring the full potential of NFV applications

On the second day of the OpenStack Summit Toby Ford, Assistant VP, IT Operations Strategic Realization at AT&T, described his personal experience with cloud offerings—both internally and externally focused. He started his talk by disclosing that when he first arrived at AT&T (as part of an acquisition) “open source was pretty much forbidden.” Today, it’s an integral part of their IT strategy.

The commitment to explore NFV is an example of their evolving strategy in action—and the company’s contribution to the OpenStack community. In fact, Ford said that the adoption of OpenStack has become a somewhat unique proxy to introduce more innovation into their DevOps process. Moreover, he sees a tremendous value in the ecosystem that has formed around this open source cloud technology.

AT&T first introduced OpenStack in their lab environment during 2010. Moving quickly, they had already progressed to production workloads by the end of 2011. Today, they have approximately 120 applications deployed on OpenStack. The platform has been deployed in seven data centers. At the end of 2014, it’s anticipated that three more data centers will join those in production.

The  virtualization use cases  have expanded from the more typical back-end applications to now include new things—such as Big Data usage scenarios. Over the next two years, AT&T plans to expand their deployment from 10 to 20 data center sites. Why so many locations? Ford says it’s all being driven by the promise of NFV—particularly within the mobility systems arena.

Why open source is now pervasive

Keeping in mind that AT&T is experiencing aggressive new competition from outside of their traditional realm, Ford acknowledged that the company’s executive leadership knows that they must change—and OpenStack is seen as a catalyst to rapidly advance a flexible and agile DevOps environment.

As an example of key areas of innovation, Ford pointed to the opportunities to transform how video entertainment content is stored and eventually delivered to AT&T U-verse customers. By going beyond prior constraints of the legacy vertically integrated infrastructure model, AT&T will be able to re-engineer their network architecture. This will allow them to distribute both key functions and multimedia content—thereby placing them much closer to the ultimate user.

Ford believes that the inherent capabilities of Cinder and Neutron will enable AT&T to have a more productive relationship with infrastructure vendors – leveraging the peer pressure within the ecosystem that naturally drives all vendors to produce high-quality code that works and scales as intended.

In addition to direct benefits, OpenStack adoption is also helping Ford and his team to propagate agile methods throughout AT&T’s vast IT organization. Besides, there has been a trickle-down effect from the early results. Now, AT&T has looked more closely at Software Defined Networking (SDN), and Ford says that he’s really excited by the opportunity to advance substantive change in the core network.

Ford believes that it’s the federation—the binding cohesion—that will ultimately make OpenStack succeed. Granted, there are recognized challenges in a number of key areas, but Ford is confident that the current path for OpenStack is both sound and stable. He anticipates more forward-thinking service providers, globally, will follow the lead of the early adopters. The positive impact is truly remarkable.

Comments are closed.

Related content.

OpenStack Tutorial – Operate Your Own Private Cloud (Full Course)

OpenStack is an open source software that provides cloud infrastructure for virtual machines, bare metal, and containers.

In this article, you will learn how to use OpenStack to operate your own private cloud.

By the end of the tutorial, you will have a core understanding of what OpenStack is and you will know the basics of setting up and administering OpenStack using the OpenMetal platform. You will also understand some commonly used OpenStack services.

In addition to creating this article version of the tutorial, I also created a video version. You can watch the video on the freeCodeCamp.org YouTube channel.

To follow along with this tutorial, it can be helpful to have a basic understanding of the Linux command line, networking, and virtualization. But none of it is required.

Thanks to OpenMetal for sponsoring this tutorial.

What is OpenStack?

OpenStack is an open source cloud computing platform that is used by organizations to manage and control large scale deployments of virtual machines, such as in a cloud computing or virtual private server environment. OpenStack is a popular choice for organizations because it is scalable, reliable, and provides a high degree of control over the underlying infrastructure.

Besides being used to manage deployments of virtual machines, OpenStack can also be used to manage storage and networking resources in a cloud environment.

In some ways OpenStack can be compared to AWS but here are some key differences between the two:

• OpenStack is an open source platform, while AWS is a proprietary platform.
• OpenStack offers more flexibility and customization options than AWS.
• OpenStack typically requires more technical expertise to set up and manage than AWS since you basically have to set up everything yourself.

Let's go into more details about what OpenStack offers.

Beyond standard infrastructure-as-a-service functionality, additional components provide orchestration, fault & service management, and other services to ensure high availability of user applications.

OpenStack is broken up into services to allow you to plug and play components depending on your needs. The OpenStack map below shows common services and how they fit together.

I won't cover every service but here is what some of the more common OpenStack services do.

Object Storage : OpenStack Object Storage (Swift) is a highly scalable, distributed object storage system.

Compute : OpenStack Compute (Nova) is a cloud computing fabric controller, which manages the allocation of compute resources.

Networking : OpenStack Networking (Neutron) is a system for managing networks and IP addresses.

Dashboard : The OpenStack Dashboard (Horizon) is a web-based interface for managing OpenStack resources.

Identity : OpenStack Identity (Keystone) is a system for managing user accounts and access control.

Image : OpenStack Image (Glance) is a service for storing and retrieving virtual machine images.

Block Storage : OpenStack Block Storage (Cinder) is a service for managing block storage devices.

Telemetry : OpenStack Telemetry (Ceilometer) is a service for collecting and storing metering data.

Orchestration : OpenStack Orchestration (Heat) is a service for orchestration and cloud formation.

Bare Metal : OpenStack Bare Metal (Ironic) is a service for provisioning and managing bare metal servers.

Data Processing : OpenStack Data Processing (Sahara) is a service for provisioning and managing Hadoop and Spark clusters.

We will be demonstrating a few of the more common OpenStack services later int his course.

There are a bunch of different ways to deploy and configure OpenStack based on the needs of your application or organization.

In this course, we will learn how to get started with OpenStack and use many of the most common features.

One of the easiest ways to get started with OpenStack is by using the OpenMetal on-demand private cloud. This allows us to quickly deploy OpenStack to the cloud and simplifies the setup process. OpenMetal provided a grant that made this tutorial possible.

While we will be using OpenMetal to learn about OpenStack, the material covered in this tutorial applies to any OpenStack deployment, not just ones that use OpenMetal. So no matter how you want to use OpenStack, this tutorial is for you.

Setting Up OpenStack on OpenMetal

To get OpenStack setup, you need to provision and set up your cloud on OpenMetal. Just follow the prompts on this OpenMetal Central page to get everything setup.

When setting up, you will have to

OpenMetal Private Clouds are deployed with OpenStack to three bare metal servers. These three servers comprise the Private Cloud Core. To OpenStack, these three servers are considered the control plane. Private Clouds are deployed with Ceph, providing your cloud with shared storage. Ceph is an open source software-defined storage solution.

Let's view the hardware assets that were created on OpenMetal. If you just created a cloud, you may already be on on the cloud management page. If not, click "manage". Now click "Assets" on the left side menu.

This page contains a list of assets included with your Private Cloud Deployment. These include your Hardware Control Plane Nodes and IP blocks for Inventory and Provider IP addresses.

The screenshot above is a list of assets in a Demo Private Cloud. Your Private Cloud can have different hardware based on the options you have selected in your deployment:

In this example, you will notice three main sections:

• 3 Cloud Core mb_small_v1 Control Plane Nodes
• Inventory IP Address Blocks
• Provider IP Address Blocks

With these Private Clouds, OpenStack is deployed with three hyper-converged control plane nodes.

You can access these Control Plane Nodes directly through SSH as the root user. This access is done through the SSH keys you provided during your Private Cloud Deployment. Use this command to connect (you will have to change the key name and IP to match your information):

ssh -i ~/.ssh/your_key_name [email protected]

Getting Started with OpenStack Horizon

Horizon is the name of the default OpenStack dashboard, which provides a web based user interface to OpenStack services. It allows a user to manage the cloud.

To access your new cloud's OpenStack dashboard (called Horizon) you will need to obtain Horizon's administrator password. The username is "admin".

To begin, SSH into one of the cloud's servers (you can use any IP address from the "Assets" page). For example:

Once you are logged in to the server, run this command:

grep keystone_admin_password /etc/kolla/passwords.yml

The password will be shown in the output as in this example:

keystone_admin_password: aB0cD1eF2gH3iJ4kL5mN6oP7qR8sT9uV

Next, launch Horizon. On OpenMetal, you can click the "Horizon" tab on the left menu.

Login using "admin" and the password you just accessed.

Create a Project in OpenStack Horizon

In OpenStack, the cloud is divided through the use of projects. Projects have associated with them users, who have differing levels of access, defined by roles. An administrator defines resource limits per project by modifying quotas.

Now we'll learn how to create a project and associate a user with it. And we will learn how project quotas can be adjusted. The Horizon interface will be similar no matter where you deploy OpenStack. This is not specific to OpenMetal.

There are three root-level tabs on the left menu in Horizon: Project, Admin, and Identity. Only users with administrative privileges can see the admin tab.

To create your first project, navigate to Identity -> Projects.

Several projects already exist, including the admin project. These projects are deployed by default and generally should not be modified.

Click the Create Project button near the top right to create a new project.

Under the Name field, specify a name for the project. This example project is called Development . You can also add Project Members and Project Groups but we are not going to cover those yet. Click Create Project to finish creating the first project.

Once created, the project appears in the Project Listing page.

While in the project listing page, you can view and adjust quotas for this project as the admin user. Quotas are limits on resources, like the number of instances.

To view the quotas for this project while in Identity -> Projects tab, find the drop down to the right with the first option being Manage Members . From this menu, click Modify Quotas to view the default quota values.

You will see a form with several tabs and you are presented with the quotas for the Compute service. Quotas exist for the Volume and Network services as well.

You may want to adjust the parameters in this form depending on your workload. Setting a value to -1 means that quota is unlimited.

How to Create a User and Associate with Project

Now that you have a project, you can associate a user with it. There is already the default admin user but now let's see how to create a new user and login with the new user.

First navigate as admin to Identity -> Users . By default, there are several users already listed, and this is expected. These are created during cloud deployment and should generally not be modified.

Click the Create User button.

On the Create User form set values for User Name , Password , Primary Project , and Role . The Email field is optional but is helpful for password resets. For the Project choose the project we created earlier.

For Role there are several options depending on the level of access required. The default OpenStack roles are reader , member , and admin . Additional roles also exist in the drop down. Reader is the least authoritative role in the hierarchy. For this example, choose member for the role.

Press Create User to create the user.

Next, log out of Horizon as admin , and log back in with your new user. Upon logging back in you are by default in the newly created project. You can see the project you are currently in at the top left and your user can be seen at the top right of Horizon.

Managing and Creating Images

Now let's learn how to upload an image (not a graphical image but a copy of a Linux installation) into OpenStack as well as create images out of an existing instance.

Images contain a bootable operating system that is used to create instances. Within your OpenMetal Cloud, there are several different images that are readily available including CentOS, Debian, Fedora, and Ubuntu. In addition to this, you have the option to upload images from other sources or create your images.

We will learn how to upload images to Glance through Horizon and how to create an image from an instance snapshot. Glance is tool for managing images that allows users to discover, retrieve, and register VM (virtual machine) images and container images. Glance uses Ceph to store images instead of the local file system.

To access images from within your Horizon Dashboard, navigate to the Projects tab. Within the projects tab, select Compute and then select Images . This tab contains a list of all your images within OpenStack.

Images can be uploaded through your Horizon dashboard by clicking the Create Image button. When creating an image you must choose the Format of the image. With our configuration, the recommended format for images is QCOW2 – QEMU emulator . QCOW2 is the most common format for Linux KVM, expands dynamically, and supports copy on write.

In order to upload an image on Horizon, you must first have the image locally on your machine. In this example, we will upload a CirrOS image. You can download a CirrOS image here .

Now click the Create Image button near the top right.

For this example we'll use the following values for the fields:

• Image Name : Name of the image
• Image Description : Optional description of the image
• File : The source file on your machine
• Format : QCOW2 – QEMU Emulator

Fill out the details as needed and submit the form. It may take some time to complete uploading the image.

Create an Instance in OpenStack Horizon

With OpenStack, instances, or virtual machines, play a large role in a cloud’s workload. OpenStack provides a way to create and manage instances with its compute service, called Nova .

Nova is the OpenStack project that provides a way to provision compute instances (aka virtual servers). Nova supports creating virtual machines, baremetal servers, and has limited support for system containers. Nova runs as a set of daemons on top of existing Linux servers to provide that service.

Now let's learn how to create an instance, including setting up a private network and router, creating a security group, and how to add an SSH key pair.

Create a Private Network

First, let's learn how to create a private network and router. Later we will create an instance on this private network. The router is created so the private network can be connected to your cloud’s public network, allowing you to assign a floating IP address to it, making the instance accessible over the Internet.

To create a private network, begin by navigating to Project -> Network -> Networks . Then click Create Network .

For this example, we'll create a network with the following details:

• Network Name : Set a name for the network. This example is called Private .
• Enable Admin State : Leave this checked to enable the network.
• Create Subnet : Leave this checked to create a subnet.
• Availability Zone Hints : Leave this option as default.

Next, move on to the Subnet tab of this form and use these details:

• Subnet Name : Set a name for the subnet. This example subnet is called private-subnet .
• Network Address : Select a private network range. For example: 192.168.0.1/24
• IP Version : Leave this as IPv4.
• Gateway IP : This is optional. If unset, a gateway IP is selected automatically.
• Disable Gateway : Leave this unchecked.

For now we will keep the default details in the Subnet Details tab.

Click Create to create the network. Once created, it appears in the list of networks.

Create a Router

You next need to create a router to bridge the connection between the private network and the public network. The public network is called External .

To create a router, begin by navigating to Project -> Network -> Routers . Click Create Router .

Input this data for this example:

• Router Name : Set a name for the router here. This example router is called Router .
• Enable Admin State : Leave this checked to enable the router.
• External Network : Choose the network External .
• Availability Zone Hints : Leave this as the default.

Once complete, create the router by pressing Create Router .

Connect Router to Private Network

Next, connect the router to the private network by attaching an interface. Performing this step allows network communication between the Private and External networks.

To attach an interface to the router, first navigate to the list of routers and locate the one previously created.

Click the name of the router to access its details page. This is where the interface is attached. There are three tabs: Overview , Interfaces , and Static Routes . To attach an interface, navigate to the Interfaces tab then load the form to attach an interface by clicking Add Interface near the top right.

On the new interface, choose the private-subnet for Subnet . If you don't set an IP address one is selected automatically. Press Submit to attach the Private network to this router. The interface is then attached and now listed.

You can visually see the network topology for your cloud by navigating to Project -> Network -> Network Topology .

The example above indicates the External network is connected to the Private network through the router called Router .

Security Groups

Security groups allow control of network traffic to and from instances. For example, port 22 can be opened for SSH for a single IP or a range of IPs.

Let's see how to create a security group for SSH access. Later we'll apply the security group we create to an instance.

To view and manage security groups, navigate to Project -> Network -> Security Groups .

You should notice a single security group called default . This security group restricts all incoming (ingress) network traffic and allows all outgoing (egress) network traffic. When an instance is created, this security group is applied by default. To allow the network traffic your instance requires, only open ports as required to just the needed IP ranges.

To create a security group for SSH, click Create Security Group near the top right.

Name the group SSH and then click Create Security Group

After creating the SSH security group, we need to add a rule allowing SSH traffic. We will allow SSH traffic from the first hardware node in this cloud to this instance.

To add a rule, load the form by navigating to Add Rule near the top right.

We'll need to obtain the IP address of the first hardware node of your cloud. You can find this using OpenMetal Central under your cloud’s Assets Page .

In the Add Rule menu, add the following information:

• Rule : Select SSH . When adding rules you can choose from predefined options. In this case, we choose the SSH rule from the first drop down.
• Description : Optional. Provide a description of the rule.
• Remote : Select CIDR .
• CIDR : Specify the IP address of your first hardware node.

Press Add to add this rule to the security group.

Create an Instance

We now have almost everything in place to create an instance.

We will need an SSH public key. An SSH public key is required to access an instance over SSH. This key is injected into the instance when created. An SSH key cannot be added to an already running instance.

We will create an instance that can be accessed over SSH from one of the cloud’s hardware nodes. So we will have to create an SSH key pair in one of the hardware nodes. The public portion of that key pair is associated with the instance we'll create soon.

** To learn how to create this key pair, see the supplementary guide: Create SSH Key Pair for an OpenStack Control Plane Node .

To create an instance, begin by navigating to Project -> Compute -> Instances . Then click the Launch Instance button.

On the details tab, fill in the following details:

• Instance Name : Set a name for the instance. This example instance is called Jumpstation .
• Description : Optional. Set a description if this applies.
• Availability Zone : Leave as the default, which is nova .
• Count : Controls the number of instances spawned. Just create 1.

Next, move to the Source tab allowing you to specify an operating system image.

Fill in the following details:

• Select Boot Source : In this example, we use Image as the boot source.
• Create New Volume : Leave this checked as Yes . This creates a new Cinder volume where the specified operating system image is copied into it. The volume ultimately exists with the Ceph cluster, in the vms pool.
• Volume Size : Allow the system to determine this for you.
• Delete Volume on Instance Delete : Leave this option set as No . If checked, when the instance is deleted, the volume is as well.
• Under the Available section, select the appropriate operating system. This example uses CentOS 8 Stream (el8-x86_64) . Clicking the up arrow will move it to the Allocated section.

This concludes configuring the instance’s source. Next, move to the Flavor tab.

Flavors are a way to define the VCPUs, RAM, and Disk space used by an instance. Pre-built flavors are available for you. For this step, select an appropriate flavor from the options under the Available heading. This example uses the m1.small flavor. Click the up arrow to move it to the Allocated section.

Next, move to the Networks tab.

In this section, you specify the network with which the instance is associated. For this example, select the Private network created previously. You can choose the External network as well, but this is generally recommended against in favor of using a floating IP should your instance require Internet connectivity.

You should only expose portions of your network as necessary. This reduces the attack surface and improves application security. If a private network is not created and an instance is created in a default cloud, it is associated with the External network. This means the instance consumes a public IP and it could be reached over the Internet.

Next, skip over the Network Ports tab and move to the Security Groups .

This is where you select security groups for the instance. This example uses the SSH security group in the Available section. Click the up arrow to move the SSH security group to the Allocated section.

As the final step, move to the Key Pair tab.

In this section, you specify an SSH public key to inject into the instance. You can upload your key at this stage using this form using the Import Key Pair button. You can also create a key pair on this tab.

We will create a key pair from the first hardware node in our cloud so this instance will be accessible over SSH from that node.

To create the SSH key pair from the first hardware node, the first step is to login to the first hardware node. You can get the IP of the hardware node from the Assets tab on OpenMetal. We already connected to this node toward the beginning of the tutorial and the command is the same. It will look something like this:

After logging in to the node, use ssh-keygen to generate an SSH key pair.

For example:

The private key is saved in the default location of /root/.ssh/id_rsa and a passphrase is set for additional security.

To view the contents of the public key, use cat /root/.ssh/id_rsa.pub .

Copy the entire key. It starts with "ssh-rsa" and continues all the way until the end.

Now back to the Key Pair tab. Click Import Key Pair.

Input the following values:

• Key Pair Name : Set a name for the SSH public key. This example public key is called jumpstation-key but it can really be anything you like.
• Key Type : This example uses an SSH Key key type.
• Public Key : Paste in the public key you just copied.

Click Import Key Pair .

Once the public key is imported, create the instance by pressing Launch Instance . (The other tabs our outside the scope of this demonstration.)

The instance goes through a build process. Allow a few minutes for this to occur. When complete, the instance appears in the Instances Listing page.

Assign and Attach Floating IP

The instance created previously is associated with a private network. Presently, the only way to access this instance is to connect to it from with the cloud’s hardware nodes. Another option for connecting is to use a floating IP. In this section, we demonstrate how to allocate a floating IP and attach it to this instance.

To allocate a floating IP, first navigate to Project -> Network -> Floating IPs . Then click Allocate IP to Project .

In the popup, make sure Pool is set to External (and optionally add a description) and then click Allocate IP to add this floating IP address for use.

In the same section, allocate the IP to the Jumpstation instance by clicking the Associate button at the far right.

Fill in the details:

• IP Address : This field comes pre-selected with the floating IP so there’s no need to change anything here.
• Port to be associated : Select the instance created previously. In this case, we use the Jumpstation instance.

Click Associate . This instance is now accessible over SSH from the first hardware node of your cloud.

To login to this instance, after you login to your hardware node, run the following command [you will have to change the IP address to the one you just associated] :

ssh -i /root/.ssh/id_rsa [email protected]

It should look something like this in your terminal:

In the next section, you will have to do something while logged in to this machine.

How to Install and Use OpenStack’s CLI

So far we've been learning how to manage OpenStack through a web browser. But it is also possible to manage through the command line using OpenStack’s CLI called OpenStackClient.

Using the command line to manage your cloud introduces more flexibility in automation tasks and can generally make an administrator’s life simpler. Let's learn how.

We'll now install the OpenStackClient on the instance we just created that we named Jumpstation.

Before installing OpenStackClient, you must obtain two files from Horizon, which are required to prepare your shell environment. Those two files are clouds.yaml and the OpenStack RC file.

• clouds.yaml : Used as a source of configuration for how to connect to a cloud
• OpenStack RC file: Used as a source of authentication for your user and project

To collect these files, log in to Horizon as your user. Navigate to Project -> API Access. Then click Download OpenStack RC File and download the OpenStack clouds.yaml and the OpenStack RC files to your machine. The files are associated with the current user and project that user is in.

Prepare and Install OpenStackClient

Next, use SSH to log in to the instance created previously. If you've been following along, then this instance can only be accessed from one of your control plane nodes. Use the instructions recently given to login to the instance.

Here are the steps to prepare and install the OpenStackClient, after you are logged in.

Step 1 : Prepare clouds.yaml and OpenStack RC files

The clouds.yaml file obtained previously must be prepared in this instance. For this demonstration, we'll save it to ~/.config/openstack/clouds.yaml . We will have to copy the contents of clouds.yaml that we downloaded to our machine from Horizon and store it as ~/.config/openstack/clouds.yaml .

Here is how you create the directory and then edit the file. Run the commands after the $ on your command line.

To get the contents from the clouds.yaml on your local computer into the instance, you will first have to open up the local version in a text editor. Then you will have to copy all the text and then paste it into the version you just created on the instance.

After you paste the text of the file into the vi editor in you terminal, use the command :wq to save and quit the editor.

The clouds.yaml file can actually be placed in several locations. For more see the Configuration Files heading of the OpenStack Victoria’s documentation.

Next, copy the contents of the OpenStack RC file from your local machine into the instance. The file can be placed anywhere so for this example we will store it in the user's home directory. The full path will be ~/Development-openrc.sh .

Create and start editing the file with the following command (don't use the $ when you copy and paste the command).

Just like before, you will have to open up the local version of the file, copy all the text, then paste it into the instance version of the file that is open in your terminal. Then, use the command :wq to save and quit the editor.

Now you have to run the file. First change the permissions:

$ chmod +x Development-openrc.sh

Then, run the file:

$ ./Development-openrc.sh

You will have to enter your OpenStack password.

Then run the command:

$ source Development-openrc.sh

Step 2 : Create a Python virtual environment

We'll create a virtual environment so we don't interfere with the system’s Python version.

In a default CentOS 8 Stream installation, the system’s Python executable is /usr/libexec/platform-python and is what will be used to create the virtual environment.

Use /usr/libexec/platform-python -m venv ~/venv to create a virtual environment in path ~/venv .

Step 3 : Activate the Python virtual environment

Use source ~/venv/bin/activate to activate the virtual environment.

After you have activated the virtual environment, the name of it will appear at the beginning of the command line. So it will look something like this:

Step 4 : Upgrade pip

Before installing OpenStackClient and to aid in a smooth installation, upgrade pip . Upgrade pip by using pip install --upgrade pip .

Step 5 : Install OpenStackClient

With everything prepared, OpenStackClient can be installed.

Note! – There exist two OpenStackClient packages: python-openstackclient and openstackclient . I recommend using python-openstackclient because it is maintained much more frequently than the prior package.

Install OpenStackClient using:

Step 6 : List servers associated with your project

For an initial command, list the servers associated with your project by running openstack server list .

Here, we can see the server created previously.

Command Structure

When using OpenStackClient, there is typically a common command pattern for what you want to accomplish. All openstack commands begin with openstack . You can execute openstack by itself to enter into a shell, where commands no longer need to be prefixed by openstack :

List all Available Subcommands

Use openstack --help to list all available subcommands. You initially see all the flags you can pass, but after scrolling a bit, the subcommand list starts:

Learn more about a Subcommand

After seeing available commands, learn more about a command by using openstack help <command> .

For example, to learn more about the openstack server command, use openstack help server :

List Items and Show Details

It is very common when using OpenStackClient to list items and the command form is typically openstack <subcommand> list . For example, openstack server list , lists all servers for the currently configured project.

Furthermore, more information about an item can be found by typically running openstack <subcommand> show <item> . For example, openstack server show Jumpstation shows the details about the instance named Jumpstation .

How Private Clouds are Deployed

Now we'll learn more about how your Private Cloud was deployed and learn more about the environment. OpenStack can be deployed in several different ways and this section highlights the characteristics of your Private Cloud. We also explain some of the advantages for this type of deployment and areas that are unique to OpenMetal.

In OpenMetal, OpenStack is containerized through Docker using Kolla Ansible. This is done through an initial deployment container called FM-Deploy. FM-Deploy provides the initial setup changes during the provisioning process of your Private Cloud. The FM-Deploy Container is a necessary part of the current architecture of your Private Cloud. The FM-Deploy Container should remain running in your Private Cloud as it is used by our systems in the event you want to add or remove nodes from your cloud.

Containerization of OpenStack

OpenMetal uses Kolla Ansible to set up Docker containers for all running services. Should you need to make any configuration changes to your nodes, Kolla Ansible should be used to push these changes. If Kolla Ansible is not used then there is a risk of these changes being reverted during any system updates.

Some advantages of containerization through docker are:

• Containers create an isolated environment reducing software dependencies
• Containers can be scaled and allow for services to balance across your cluster
• Containers provide increased flexibility for test releases, patches, and automation
• Containers have a consistent and repeatable deployment and a shorter initialization time

Disk Storage and Ceph

In OpenMetal, disk storage is provided through Ceph. Ceph is an object storage interface that can provide interfaces for multiple different storage types on a single cluster. In OpenMetal Ceph is comprised of two elements: object storage and block storage.

Ceph object storage utilizes Ceph Object Storage Gateway daemon (RADOSGW). With OpenMetal clouds, Ceph’s RGW replaces Swift so there is no Docker container for Swift. Instead, Swift endpoints are connected directly to the RGW. Authentication for RGW is handled through Keystone in /etc/ceph/ceph.conf .

Ceph block storage connects to the Cinder service utilizing Ceph’s RADOS Block Device. Within your cloud, those objects are stored in Ceph pools. Ceph provides a layer of abstraction that allows objects to be recognized as blocks.

Some advantages of using ceph are:

• Data is self-healing and will redistribute data across your cluster in the event of power, hardware, or connectivity issues
• Data is replicated and highly available
• Ceph has the ability to run on commodity hardware and to mix hardware from different vendors

Introduction to Ceph

Ceph is an open-source, distributed storage system that provides object, block and file storage interfaces from a single cluster.

Ceph was selected as the storage solution for Private Cloud Core OpenStack clouds due to its ability store data in a replicated fashion. The data stored in the Ceph cluster is accessible from any of your cloud’s control plane nodes. The storage is considered shared across all nodes, which can make recovering an instance and its data trivial.

Let's learn how to check the status of your Ceph cluster and see available disk usage using the command line.

Check Ceph Status

First, make sure you are logged into one of your cloud’s control plane nodes (not an instance). To check the status of your Ceph cluster, use ceph status .

Check Ceph Disk Usage

To check the available disk space in your Ceph cluster, use ceph df .

Maintaining OpenStack Software Updates

Software in the OpenStack ecosystem evolves over time, either through new feature additions, bug fixes, or when vulnerabilities are patched. Part of operating an OpenStack cloud involves maintaining its software through updates. In this section, we point out the sections in an OpenMetal cloud where software updates occur and explain best practices when performing updates.

The software of an OpenMetal cloud that can be updated include each hardware node’s package manager and the Kolla Ansible Docker images. Ceph updates are handled through the node’s package manager.

Now we will specifically cover the step to perform package manager updates.

• Migrate Workload

Package manager updates requiring a server reboot to an OpenMetal control plane node can be disruptive to any workload running on it. Prior to performing disruptive actions, it may be possible to migrate instances another node running the Compute service. For information on how to migrate instances, see OpenStack Nova’s documentation .

2. Update One Node at a Time

While performing package manager updates, ensure updates occur successfully for one hardware node before updating another node.

3. Disable Docker

Before updating the package manager, ensure the Docker socket and service within SystemD are stopped and disabled. For example:

4. Upgrade Host OS Packages

After verifying the Docker socket and service are stopped, perform the package manager updates.:

5. Determine Reboot Need

Once package manager completes, check if a reboot is required with dnf-utils needs-restarting and the reboot hint flag (-r):

6. Ceph Maintenance

This step is optional and only required if the node needs to be rebooted.

Prior to reboot, if the node is part of the Ceph cluster automatic OSD removal and data rebalance should be temporarily suspended. To do so, perform:

This will reduce rebuild time and help ensure the node rejoins the cluster automatically.

Once the node reboots and a healthy Ceph cluster is confirmed, these parameters must be unset. To unset this configuration, perform:

Reboot if Required

Reboot the node if required:

You may have to wait a minute or two before you can log back in to the control plan node.

Verify Successful Reboot

When the node comes back online, SSH into it to verify the OpenStack Docker containers have started. Additionally, if this node was part of the Ceph cluster, check Ceph’s cluster status.

To verify the Docker containers have started, use docker ps . You should see a number of Docker containers running. Under the STATUS column, each container should reflect the status Up .

Next, if this node is part of a Ceph cluster, check Ceph’s status using ceph status .

The above output shows the status as HEALTH_OK , indicating the Ceph cluster is healthy. Ceph is naturally resilient and should recover from a node being rebooted.

View your Private Cloud’s Resource Usage

Now let's look at how to find the resource usage of your private cloud. We will explore how to utilize the Horizon Dashboard to determine the total memory and compute usage for a project as well as how to view instances stored on each node. Next, we'll look at disk usage by explaining how to briefly interact with your cloud’s Ceph cluster using the command line. Finally, we'll go over adding and removing nodes from your Ceph cluster.

There are currently three variations to private cloud deployments: Small, Standard, and Large. All private cloud deployments have a cluster of three hyper-converged servers but will have different allocations of memory, storage, and CPU processing power depending on the configuration and hardware. In addition, you have the option of adding additional hardware nodes to your cluster.

View Memory and Compute Usage in Horizon

To view the resources used by your cloud, you have to be the admin user and assigned to the admin project. Once you are in the admin project, navigate to Admin -> Compute -> Hypervisors . This section, lists the following items:

• Memory Usage
• Local disk usage

View Instance State Across Cluster

There is also an option to see the location of your instances within your cluster. To view this information, navigate to Admin -> Compute -> Instances . You have the option to see the project, the host, as well as the IP address, and state.

How to Access Resource Information from Ceph

To access information regarding your Ceph cluster’s resource pools, you will need to use Ceph’s CLI. These are a summary of some useful resource monitoring and health commands.

• ceph -s to check the status of Ceph
• ceph df to list the disk usage overview
• ceph health detail provides details about existing health issues
• ceph osd pool ls to list pools add detail for additional details regarding replication and health metrics

Create and Restore Volume Backups

With Private Clouds, you have the ability to create backups and snapshots of your volume data. If a volume’s data goes corrupt or is removed by mistake, having a copy of that data could be invaluable. Now we'll look at how to create and recover volume backups using Horizon.

First of all, let's review what a volume is. Volumes are block storage devices that you attach to instances to enable persistent storage. You can attach a volume to a running instance or detach a volume and attach it to another instance at any time. You can also create a snapshot from or delete a volume.

Create a Volume Backup

Navigate in Horizon to Project -> Volume -> Volumes.

Create a backup of your volume by selecting from the drop down Create Backup .

You will need to fill in the following fields.

• Backup Name : Specify a name for the volume backup
• Description : Provide a description if necessary
• Container Name : Leave this blank otherwise the volume backup cannot be created. Horizon tells you if this field is blank, the backup is stored in a container called volumebackups , but this is not the case with our configuration. With Private Clouds, all volume backups created this way are stored in the Ceph pool called backups .
• Backup Snapshot : If applicable specify a snapshot to create a backup from

After submitting the form, you are navigated to Project -> Volume -> Volume Backups where you can see the volume you just created a backup of.

Test Volume Backups

Creating backup copies of your important data is only one part of having a solid backup and recovery plan. Additionally, consider testing backed-up data to ensure if something unexpected does happen that restoring your backups will actually be useful. To test volume backups, you can restore a volume backup within OpenStack alongside the original volume and compare the contents.

Restore a Volume Backup

To restore a volume backup, being by navigating in Horizon to Project -> Volume -> Volume Backups .

Next, find the volume backup you wish to restore and from its drop down on the right, select Restore Backup .

Choose the volume to restore to, or have the system restore the backup to a new volume.

Ceph, Volumes, and Data Durability

When volume backups are created, they are stored in your cloud’s Ceph cluster in a pool called backups. By default, the Ceph cluster is configured with host level replication across each of your cloud’s three control plane nodes. With this configuration, your cloud could suffer losing all but one Ceph node and still retain all of the cluster’s data.

You should now have a basic understanding of OpenStack and some of the things you can do with it. Good luck setting up your own OpenStack system.

I'm a teacher and developer with freeCodeCamp.org. I run the freeCodeCamp.org YouTube channel.

If you read this far, thank the author to show them you care. Say Thanks

Learn to code for free. freeCodeCamp's open source curriculum has helped more than 40,000 people get jobs as developers. Get started

• Workshop Organizers
• Registration Information
• Registration Discounts
• At a Glance
• Workshop Program
• Live Streaming
• Birds-of-a-Feather Sessions
• Sponsorship
• Hotel and Travel Information
• Help Promote!
• For Participants
• Call for Papers
• Past Workshops

usenix conference policies

• Event Code of Conduct
• Conference Network Policy
• Statement on Environmental Responsibility Policy

You are here

Connect with us, dissecting open source cloud evolution: an openstack case study.

Salman A. Baset, Chunqiang Tang, Byung Chul Tak, and Long Wang, IBM T.J. Watson Research Center

Open source cloud platforms are playing an increasingly significant role in cloud computing. These systems have been undergoing rapid development cycles. As an example, OpenStack has grown approximately 10 times in code size since its inception two and a half years ago. Confronting such fast-pace changes, cloud providers are challenged to understand OpenStack’s up-to-date behaviors and adapt and optimize their provisioned services and configurations to the platform changes quickly. In this work, we use a black-box technique for conducting a deep analysis of four versions of OpenStack. This is the first study in the literature that tracks the evolution of a popular open source cloud platform. Our analysis results reveal important trends of SQL queries in OpenStack, help identify precise points for targeted error injection, and point out potential ways to improve performance (e.g. by changing authentication to PKI). The OpenStack case study in this work effectively demonstrates that our automated black-box methodology aids quick understanding of platform evolution and is critical for effective and rapid consumption of an open-source cloud platform.

Salman A. Baset, IBM T. J. Watson Research Center

Chunqiang tang, ibm t. j. watson research center, byung chul tak, ibm t. j. watson research center, long wang, ibm t. j. watson research center, open access media.

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.

Presentation Video 

Presentation audio.

Download Audio

• Log in  or     Register to post comments

Gold Sponsors

Silver Sponsors

Bronze Sponsors

Media Sponsors & Industry Partners

• Privacy Policy

Storware | An OpenStack Case Study

Learn more about how Storware provides backups, the challenges they have helped resolve and more.

Having backup and recovery software is crucial for data safety, and Storware has created a way to perform agentless backups for OpenStack resources. Ranging from virtual machines (VMs), volumes and other critical data, their software can provide backups for agents on each individual instance.

In particular, Storware provides its services to companies like Cloudfire, an Italian cloud provider that specializes in advancing IT innovation and automation for System Integrators and Telco service providers. With the assistance of Storware, they were able to overcome several challenges related to the backup and recovery process. Overall, this software has helped enhance security and operational excellence within their OpenStack environment. Learn more about how Storware provides backups, the challenges they have helped resolve and the compatibility of their software with other open source technologies.

How does Storware provide backups for OpenStack environments?

Storware Backup & Recovery supports OpenStack backups in the following ways:

• Disk attachment through Cinder with changed block tracking, with support for all hypervisors and storage solutions.
• Disk image transfer, specifically for KVM hypervisors with VMs using QCOW2.
• Backup support for Ceph-based volumes, with incremental backup capabilities.
• Last but not least, Storware provides an OpenStack Horizon integration plugin, allowing for direct management within the OpenStack dashboard.

What challenges in deploying in OpenStack does Storware help fix?

OpenStack is highly flexible, enabling users to combine multiple components and create unique environments. Some might consider OpenStack to be a complex platform, and deploying backup and recovery solutions can indeed pose challenges. However, Storware simplifies backup processes and offers a user-friendly interface for administrators to manage backup tasks, thanks to its solid integration capabilities.

Due to the versatile applications and varied use cases of OpenStack environments, scalability is crucial for backup and recovery solutions. Storware is designed to be scalable and capable of supporting a wide range of OpenStack environments, from small and medium-sized businesses to large enterprises, without impacting performance.

Furthermore, Storware contributes to resource optimization within OpenStack by providing features such as deduplication and compression, which reduce the storage footprint required for backups.

Moreover, all backup and recovery jobs can be easily managed and automated from a single dashboard. Additionally, Storware aids in compliance efforts by offering backup policies and retention settings that align with regulatory requirements for data protection and retention.

What other open source technologies are compatible with Storware’s backup services?

Storware’s backup services are designed to be versatile and compatible with various open-source technologies:

• Deployment with a focus on security: Storware supports Linux-based installations on popular distributions like CentOS, SUSE, Rocky, AlmaLinux, Oracle and Red Hat.
• KVM and KVM-based environments: Storware provides backup solutions for KVM virtualization, often used in conjunction with OpenStack.
• Ceph: Storware supports backup and recovery of data stored in Ceph RBD storage clusters, commonly integrated with OpenStack for scalable and distributed storage.
• Docker and Kubernetes: Storware supports backup and recovery for containerized applications and Kubernetes clusters.
• Database Systems: Storware offers backup and recovery solutions for open-source databases like MySQL, PostgreSQL or MongoDB, commonly used in various applications and environments.

Share your own use case by filling out the  OpenStack User Survey !

Tags: backup , case studies , cloud , Linux , OpenStack , recovery , Use Case , VMs

Recent Posts

• Meet OpenInfra Europe Advisory Board Member: Kurt Garloff - December 21, 2023
• NVIDIA’s Journey with Kata Containers and Confidential Computing | OpenInfra Live Recap - December 18, 2023
• Large Scale Ops Deep Dive: Samsung SDS | OpenInfra Live Recap - December 14, 2023

Inside Open Infrastructure: October 2023 PTG Recap

PTG Recap, project updates and the latest news from the OpenInfra community!

OpenStack Security: A Practical Guide

Here are a few severe vulnerabilities discovered in OpenStack over the past few years, codified as common vulnerability enumerations (CVEs).

Running Relax-and-Recover to save your OpenStack deployment

Carlos Camacho of Red Hat walks you through this disaster recovery solution for Linux.

Examination of Cost Differences Between Private Cloud and Hyperscalers | OpenInfra Live Recap

Georgia Cyber Center | An OpenStack Case Study

Appreciating the Need for Edge Location Standards or Schemas

Software Deployment Options in OpenStack: A Quick Guide

• Tacker’s Contributions to ETSI-NFV Standards
• Virtual Labs | An OpenStack Case Study
• Inside Open Infrastructure: January 2024
• OpenStack Neutron – Networking In Cloud Demystified
• 10 Reasons to Back Upstream Open Source Contributions
• Open Infrastructure Events
• Superuser Award
• Superuser TV
• Uncategorized
• User Stories

Intrusion Detection and Prevention in OpenStack: A Case Study on Enhancing Security and Threat Detection

• Original Research
• Published: 28 October 2023
• Volume 4 , article number  830 , ( 2023 )

Cite this article

• Sanjay Adiwal   ORCID: orcid.org/0000-0001-8350-0999 1 &
• Mohammad Misbahuddin 2  

69 Accesses

Explore all metrics

Internet usage has experienced exponential growth over the past two decades. Nowadays, numerous human activities heavily rely on the Internet and information technology. The Internet is utilized for various purposes, including communication, online financial transactions, banking, entertainment, work, and educational tasks. As our dependency on the Internet increases, a substantial amount of data and sensitive information is inevitably transmitted over the Internet, especially through cloud services. Consequently, malicious attackers are actively seeking security loopholes and vulnerabilities, specifically targeting cloud environments. Intrusion Detection and Prevention System (IDPS) is a hardware appliance or a software application that monitors a network for malicious activity or breaches of policies, and based on the rules created, blocks any potential malicious activity. Cloud service providers and users now require IDPS as part of their security infrastructure. This paper investigates how an open-source IDPS named Snort can be effective in OpenStack-based cloud environments. The effectiveness of the Snort-based cloud security solution is demonstrated through a series of experiments.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

A Collaborative Study of Intrusion Detection and Prevention Techniques in Cloud Computing

HIDSC2: Host-Based Intrusion Detection System in Cloud Computing

Intrusion Detection System with Snort in Cloud Computing: Advanced IDS

Data availability.

The data that support the findings of this study are available on request from the corresponding author.

Internet Stats & Facts, List of Internet, eCommerce, Hosting, Mobile & Social Media Statistics for 2021.2023. https://websitesetup.org/news/internet-facts-stats . Accessed 20 Mar 2023.

Kumar J, Bindhumadhava BS, Santhanavijayan A, Rajendran B. An adaptive neural network for email spam classification. In: Fifteenth International Conference on Information Processing (ICINPRO). 2019; https://doi.org/10.1109/ICInPro47689.2019.9092278 .

Gupta A, Lalit S. Detecting attacks in high-speed networks: issues and solutions. Inform Secur J Glob Perspect. 2020;29:5–6.

Google Scholar  

Kumar J, Santhanavijayan A, Rajendran B. Cross site scripting attacks classification using convolutional neural network. In: 2022 International Conference on Computer Communication and Informatics (ICCCI). 2022; https://doi.org/10.1109/ICCCI54379.2022.9740836 .

Khalil IM, Abdallah K, Muhammad A. Cloud computing security: a survey. Computers. 2014. https://doi.org/10.3390/computers3010001 .

Article   Google Scholar  

Carlin S, Kevin C. Cloud Computing security. In: Pervasive and ubiquitous technology innovations for ambient intelligence environments. IGI Global; 2013.

Dimitrios Z, Dimitrios L. Addressing cloud computing security issues. Futur Gener Comput Syst. 2012. https://doi.org/10.1016/j.future.2010.12.006 .

A practical guide to zero trust implementation, roadmap: the zero trust security playbook. 2023. https://www.forrester.com/report/A+Practical+Guide+To+A+Zero+Trust+Implementation/-/E-RES157736 , Accessed 20 Mar 2023.

Zhou Z, Zhongwen C, Tiecheng Z, Xiaohui . The study on network intrusion detection system of Snort. In: International Conference on Networking and Digital Society. 2010.

Snort System software. Snort—a free open source network intrusion detection system.2023. https://www.Snort.org/ . Accessed 10 May 2023.

Sanjay A, Rajendran B, Shetty PD. DNS amplification & DNS tunneling attacks simulation, detection and mitigation approaches. In: International Conference on Inventive Computation Technologies (ICICT). 2020; https://doi.org/10.1109/ICICT48043.2020.9112413

Download references

Author information

Authors and affiliations.

RISE, C-DAC, Bengaluru, India

Sanjay Adiwal

ACTS & BD, C-DAC, Bengaluru, India

Mohammad Misbahuddin

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Sanjay Adiwal .

Ethics declarations

Conflict of interest.

The authors have no relevant financial or non-financial interests to disclose. The authors have no conflicts of interest to declare that are relevant to the content of this article. All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript. The authors have no financial or proprietary interests in any material discussed in this article.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This article is part of the topical collection “Advances in Internet Research and Engineering 2023” guest edited by Sudarsan S D, Mohit Sethi and Balaji Rajendran.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Adiwal, S., Misbahuddin, M. Intrusion Detection and Prevention in OpenStack: A Case Study on Enhancing Security and Threat Detection. SN COMPUT. SCI. 4 , 830 (2023). https://doi.org/10.1007/s42979-023-02265-3

Download citation

Received : 06 July 2023

Accepted : 19 August 2023

Published : 28 October 2023

DOI : https://doi.org/10.1007/s42979-023-02265-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Cloud security

Advertisement

• Find a journal
• Publish with us
• Track your research

Advertise with us

• ARTIFICIAL INTELLIGENCE
• AUDIOVISUAL
• AUGMENTED & VIRTUAL REALITY
• BUSINESS INTELLIGENCE
• CLOUD MIGRATION
• DATA ANALYTICS
• DATA INTEGRATION
• DATA VISUALIZATION
• DIGITAL ACCESSIBILITY
• DIGITAL TRANSFORMATION
• DIGITAL TWIN
• E-WASTE MANAGEMENT
• EDGE COMPUTING
• ENTERPRISE ARCHITECTURE
• GAMING TECH
• HIGH PERFORMANCE COMPUTING
• INFRASTRUCTURE SECURITY
• INSIGHT ENGINES
• INTERNET OF THINGS
• MOBILE APPLICATION
• PREDICTIVE ANALYTICS
• QUANTUM COMPUTING
• REMOTE TECH
• SOFTWARE TESTING
• VIRTUAL DATA ROOM
• WEB DEVELOPMENT
• BANKING & INSURANCE
• CAPITAL MARKETS
• CONSTRUCTION
• CONTACT CENTER
• FOOD AND BEVERAGES
• LAW ENFORCEMENT
• MANUFACTURING
• MEDIA & ENTERTAINMENT
• PHARMA & LIFE SCIENCES
• PUBLIC SECTOR
• SEMICONDUCTOR
• TRAVEL & HOSPITALITY
• BUSINESS PROCESS MANAGEMENT
• CLOUD COST MANAGEMENT
• CLOUD COST OPTIMIZATION PLATFORM
• COLLABORATION
• CONVERSATIONAL
• CUSTOMER ENGAGEMENT
• CUSTOMER EXPERIENCE MANAGEMENT
• CUSTOMER RELATIONSHIP MANAGEMENT
• CYBER SECURITY
• DATA CENTER
• DATA SECURITY
• DIGITAL ASSET MANAGEMENT
• DIGITAL EXPERIENCE
• DIGITAL RIGHTS MANAGEMENT
• DIGITAL SIGNAGE
• DISASTER RECOVERY
• DOCUMENT MANAGEMENT
• DOCUMENT PROCESSING
• END-TO-END-ENCRYPTION
• ENTERPRISE APPLICATION INTEGRATION
• ENTERPRISE ASSET MANAGEMENT
• ENTERPRISE CONTENT MANAGEMENT
• ENTERPRISE DATA MANAGEMENT
• ENTERPRISE INFORMATION MANAGEMENT
• ENTERPRISE MOBILITY
• ENTERPRISE PERFORMANCE MANAGEMENT
• ENTERPRISE RESOURCE PLANNING
• ENTERPRISE RISK MANAGEMENT
• FACILITY MANAGEMENT
• FIELD SERVICE
• GAMIFICATION
• GEOGRAPHIC INFORMATION SYSTEM
• IDENTITY AND ACCESS MANAGEMENT
• IDENTITY GOVERNANCE AND ADMINISTRATION
• INTELLIGENT DATA CAPTURE
• IT SERVICE MANAGEMENT
• IT SERVICES
• KNOWLEDGE MANAGEMENT
• LEGACY MIGRATION AND MODERNIZATION
• LOW CODE PLATFORM
• MANAGED COMMUNICATIONS
• MANAGED IT SERVICES
• MANAGED PRINTING
• MOBILE PAYMENT
• PAYMENT AND CARD
• PRESCRIPTIVE ANALYTICS
• PROCESS ORCHESTRATION
• PRODUCT LIFECYCLE MANAGEMENT
• PRODUCT MANAGEMENT
• PROFESSIONAL SERVICES AUTOMATION
• PROJECT MANAGEMENT
• QUALITY MANAGEMENT
• RISK ANALYTICS
• SCHEDULING SOFTWARE
• SMART INFRASTRUCTURE
• SUBSCRIPTION AND BILLING
• UNIFIED COMMUNICATIONS
• VIDEO SOLUTIONS
• CONTRACT MANAGEMENT
• DIGITAL MARKETING
• HUMAN RESOURCE
• PROCUREMENT
• SUPPLY CHAIN
• Whitepapers
• Conferences

Creating a long-term blueprint for the solution and the deployment plan

Amax Information Technologies

How to provide telecommunications as well as IT, to improve internet connections

CloudBerry Lab

How customers and end users can access personal portals

Memset Hosting Limited

How to rebuild the development and deployment environment

Apriorit Inc

Read more to know about the AWS cloud migration

Spark Support Infotech Pvt Ltd

ON THE DECK

OpenStack 2017

Top Vendors

OpenStack 2016

OpenStack 2015

I agree We use cookies on this website to enhance your user experience. By clicking any link on this page you are giving your consent for us to set cookies. More info

However, if you would like to share the information in this article, you may use the link below:

https://openstack.cioreview.com/casestudy/

NEW! Mirantis Academy -   Learn confidently with expert guidance and On-demand content.   Learn More

PUBLIC SECTOR

Société Générale

Mirantis Supports Société Générale’s Journey of Cloud Native Transformation

Read Case Study

Read our latest customer stories

Netsons Expands Cloud Computing Services Across Europe Using Mirantis OpenStack Private Clouds

Netskope Expands Secure Access Service Edge with Mirantis OpenStack Private Cloud

Mirantis Helps Inmarsat Build Its Next-Generation Communications Network

Vaudoise Insurance Builds Out Modern Microservices with Mirantis Kubernetes Engine

Containers | Financial

Mirantis helps Asian bank scale mobile banking for millions of customers during COVID-19 pandemic

Containers | Business Services

Docker Enterprise Delivers Security at Scale for ADP

Containers | Defense

Aerospace Leader

Mirantis Kubernetes Engine gives Aerospace company an on-prem container platform that drives a secure software supply chain–with global impact.

Alm. Brand Saves 90 percent of Maintenance Time in 10 Months

OpenStack | Telecom

Asian Telecom Operator

Leading telecommunications operator uses Mirantis Cloud Platform to quickly scale for 100 million new customers in less than 6 months

OpenStack | Business Services

Bailian Group

Chinese retail giant launches massively scalable omni-channel e-commerce and IaaS platform

Containers | Oil & Gas

Baker Huges GE

Baker Hughes GE Speeds Application Deployment Times from 6+ Months to 10 Minutes

BCG Gamma Transforms Analytics and Machine Learning with Docker Enterprise

Containers | Travel

Carnival Corporation

Carnival Corp Leverages Docker Enterprise to Create Unique Guest Experiences on Its Cruise Ships

Citizens Bank

Citizens Bank Modernizes the Mortgage Experience and Cuts Infrastructure Costs with Docker Enterprise

CloudVPS uses OpenStack to deliver a public cloud.

Containers | Manufacturing

Complevo Drives Automotive Leaders Forward with Mirantis Container Platforms

Containers | Technology

Conoa collaborates with businesses for successful transformations with open source

Containers | Education

Cornell Lab of Ornithology

The Cornell Lab of Ornithology accelerates development on high-availability apps for a global user-base.

Cornell University

Cornell University Accelerates Cloud Migration

Defense Agency Cybersecurity Office

Defense Agency Cybersecurity Office focuses on its mission, turning to Mirantis for the only STIG-ready managed container service on the market.

OpenStack | Financial

European Banking Leader

Global 500 bank deploys OpenStack private cloud to speed product introduction and reduce costs

Fortune 500 Credit Card Company

Leading credit card company speeds time-to-market and lowers platform costs

OpenStack | Media

Fortune 500 Pay-TV Provider

Leading satellite operator uses Mirantis Cloud Platform to launch an over-the-top streaming service in less than 2 months

Containers | Pharma

GlaxoSmithKline

GlaxoSmithKline Accelerates Drug Discovery with Docker Enterprise

OpenStack | Healthcare

European Medical Center

European Medical Center Drives Cutting-Edge Research with Mirantis OpenStack Private Cloud

Global Pharma

How a global pharma company rapidly migrated 3,000 Kubernetes environments to Amazon EKS and cut application delivery costs by 60%

Halliburton

Application Portability to the Cloud and Edge

InterSystems

InterSystems Enhances and Accelerates its Use of Kubernetes with Training from Mirantis

Jabil Modernizes Legacy Infrastructure from the Datacenter to the Edge with Docker Enterprise

Japanese Telecom

Leading operator deploys Kubernetes and OpenStack cloud for cloud-native enterprise apps

Jireh Modernizes 15+ Year Old HP-UX Manufacturing Apps with Docker Enterprise

Containers | Government

Kadaster Builds a Flexible Platform to Handle 8 Billion Transactions a Year

Leading Telecom

With OpenStack, leading Asia Pacific telecom reduces provisioning time from weeks to seconds

Liberty Mutual

Agility in the Cloud and Datacenter for Developers at Liberty Mutual with Docker Enterprise

Lindsay Corporation

Lindsay Corporation Increases ROI and Modernizes Agriculture with Docker Enterprise & Microsoft Azure

MetLife Transforms Customer Experience with Application Modernization

Containers | Healthcare

Mutuelle Just

Mirantis & Mutuelle JUST - Building Container Infra 'Just Right'

OpenStack | Education

Naturalis Biodiversity Center

Scientific institute enables faster research calculations and resource scaling with OpenStack

OpenStack | Technology

Netsons Expands Cloud Computing Services Across Europe with OpenStack Private Clouds

PayPal Manages 200,000 Containers in the Cloud to Speed Transactions

Pharma Giant

Pharma giant boosts innovation with hybrid, multi-cloud platform

Containers | Food Products

Premium Food Supplier

Premium food supplier serves up faster software delivery with Mirantis Kubernetes Engine and Lens.

ProSiebenSat.1 Media

European media giant taps OpenStack-AWS hybrid cloud to accelerate time to market

Protector Forsikring

Mirantis Helps Protector Forsikring Challenge Insurance Industry with Software Agility and Innovation

Royal Bank of Canada

RBC Capital Markets Expands Risk Analysis Models for Data Scientists by Scaling to GPUs

MultiCloud | Financial

S&P Global Centrally Manages On-prem and Public Cloud Kubernetes Clusters.

Seagate Factories Improve Product Quality with AIML at the Edge

SEB Gives Developers Fast Deployments with CI/CD

Shenzhen Stock Exchange

Chinese stock exchange increases performance of IT service offerings, improves elasticity and enables growth

Splunk Improves Product Quality with Docker Enterprise

Largest Middle East and Africa telecom provider builds innovative public OpenStack cloud

Containers | Insurance

OpenStack | Manufacturing

Volkswagen drives innovation across business and consumer applications with agile infrastructure from Mirantis

VR Group, Finnish Railways

Finnish Railways Modernizes Key Transportation Apps with Docker Enterprise

Customer Reviews

900 E Hamilton Avenue Suite 650 Campbell, CA 95008 +1-650-963-9828

Privacy Policy

Do Not Sell My Information

UK Modern Slavery Act Statement

• What is ZeroOps
• ZeroOps for developers
• ZeroOps for cloud operators
• Modern Application Delivery
• Cloud Platform Operations
• Lens Desktop
• amazee.io's Lagoon
• Mirantis Container Cloud
• Mirantis Kubernetes Engine
• Mirantis OpenStack for Kubernetes
• Mirantis Container Runtime
• Mirantis Secure Registry
• Why Mirantis
• Customer Stories
• Open Source
• Sustainability

© 2005 - 2024 Mirantis, Inc. All rights reserved. “Mirantis” and “FUEL” are registered trademarks of Mirantis, Inc. All other trademarks are the property of their respective owners.

The State of Pricing, Demo, and Case Study Pages

When you think about B2B SaaS websites, regardless of how disruptive the product is, there are always some “sacred pages” that most companies maintain. These pages are so ingrained in the fabric of the site that their existence is rarely, if ever, questioned.

Let’s think about the pricing pages; if the pricing isn’t transparent, does a company really need a pricing page? If so, is saying “submit the form to find out about pricing” enough? This was the initial question. The next question was about case studies. Many companies aim to feature as many case studies as possible on their websites, but to what end? Do website visitors actually engage with these pages? Do these pages influence conversions? Do companies even track these pages? These realizations sparked our curiosity; unsure whether these were merely following the status quo or proven best practices, we set out to myth-bust these.

Methodology

MQL: High-intent demo, pricing page, contact us submissions. Basically every hand-raiser on the website. Ebook form submissions, lead-gen stuff, webinar registrations were not counted as MQLs. In the first Labs report, some people got a bit heated with us for saying MQLs, but I won’t be changing this simply because I like the sound of it.

Demo Pages: Any page that says “request a demo”, “book a call”, “contact sales” etc. including PPC landing pages. 

Pricing Pages: Any page that says “pricing”, “quote” etc.

Sample Size: Over 31M unique website visitors from 80 companies that have both pricing and demo pages on their websites, all B2B SaaS. 

37.5% of the dataset had non-transparent pricing compared to 62.5% with transparent pricing.

Sample Description: From $5M ARR to $1B ARR; average ACV from $5K to $120K. 

Caveats: Since we’re not looking at the entire funnel, we haven’t had a specific time period filter for this analysis, meaning that if a new customer implemented HockeyStack in December, we included their data for January (not February to not miss any pipeline data). So, some companies might have a month-long data, while others might have 18-month-long data.. 

TLDR – for B2B SaaS companies: 

Bounce Rate: 59%

Avg. Session Duration: 2 minutes 35 seconds

Page per Session: 2.62

Part I – Transparent vs. Non-Transparent Pricing

Let’s start with the website metrics. Our dataset shows that on average, the bounce rate for pricing pages is 39%, which is significantly lower than the average website bounce rate. We see an average of 35% bounce rate when there’s no transparent pricing, and 42% when there’s transparent pricing.

I think the reason for this could be because if there’s transparent pricing and that pricing doesn’t align with the buyer’s budget, they bounce right away; but if there’s no pricing, they stick around more, perhaps to discover or estimate the potential cost. For example, if they don’t see pricing, then these visitors might then be looking for customer logos or case studies – since if there are enterprise logos, one can safely assume that the cost will be higher. We need to look at additional metrics to make more assumptions.

When pricing is not transparent, website visitors visit an average of 2.57 pages compared to 4.26 pages when pricing is transparent. This suggests that transparent pricing might have a higher bounce rate, but if visitors don’t bounce, they end up visiting 65% more pages and spend an average of 3 minutes and 25 seconds on the website compared to 2 minutes and 31 seconds when pricing is not transparent. 

So, here what we are seeing is that when there’s no transparent pricing, the bounce rate is lower; however, website visitors end up visiting fewer pages and spend less time. As suggested above, maybe they just try to estimate the pricing by looking at other pages.

Then what do we see on the conversion side? On average, pricing pages have a 3.8% conversion rate. When there’s transparent pricing, this conversion rate drops to 2.8%; and when there’s no transparent pricing, this conversion rate jumps to 4.6%. I believe the main reason for this is curiosity; lower conversion rates in transparent pricing pages indicates that visitors only convert if they know they can afford the tool, and higher conversion rates in non-transparent pricing pages indicate that visitors might just want to learn about the initial cost. 

It’s rare—if any—that B2B companies actually differentiate between pricing and demo calls. Thinking of the origins of having forms on non-transparent pricing pages, it’s possible that pricing pages might seem more approachable, more relaxed/chill in a way, hence tempting buyers who aren’t fully ready to submit the pricing form over a demo request. But if they’re not fully ready, are they truly the conversions you want to get?

Let’s look at the pipeline data now. Non-transparent pricing pages have an average of 10.31% submit:pipeline conversion rate; for transparent pricing pages, this rate is 17.50%. 

Although non-transparent pricing pages have better form submission rates (4.6% vs. 2.8%) and they -on paper- generate more MQLs; on the pipeline side what we’re seeing is that transparent pricing page MQLs convert into pipeline 1.7x better.

This is very interesting because when we look at the form submission rate differences, non-transparent pages have a 1.64x better conversion rate; while on the pipeline conversion rate side, transparent pages have a 1.7x better conversion rate. This data shows that transparent pricing pages generate a better pipeline than non-transparent pages.

What about submitting demos? How does transparent pricing forms impact demo form submissions compared to non-transparent pricing? For this, we’ll use Lift Reports where we can easily zoom in on specific marketing activities to measure the incremental contribution by using control and treatment groups.

When we use the lift analysis, our dataset shows that when there’s non-transparent pricing, users are 9.5% less likely to submit the demo form compared to transparent pricing; the control group difference is 0.92 vs. 0.84x.

Hence, not only do transparent pricing pages bring better a pipeline but also they influence the demo form submission rates better than the non-transparent pricing ones.

Part II – Pricing Pages vs. Demo Pages

I’ve started including “if you have any ideas, please let me know” at the end of each report. Surprisingly, five people brought up the same question: what are the conversion benchmarks for pricing vs. demo pages?

From my experience with previous clients and companies, although the metrics varied, pricing pages consistently showed lower conversion rates than demo pages. This discrepancy was generally accepted, underpinned by the belief that if someone doesn’t convert from the demo, they might still convert from the pricing page—effectively “doubling” the chance for a conversion.

Let’s start with bounce rates. While the average bounce rate for pricing pages was 39%, for demo pages it jumps to 70%. 

At first glance, the significant difference between the bounce rates of pricing and demo pages might seem alarming, but it’s important to consider other factors here.

My first hypothesis was that maybe marketers direct most of their paid traffic to the demo pages regardless of whether it’s high-intent traffic or a cold audience, and this was causing the high bounce rate. I refuted this hypothesis by looking at the total unique users difference. According to our dataset, pricing pages have 13x more unique visitors than the demo pages, so it’s not anything like “demo pages get lots of visitors and these visitors bounce.” This situation left me with three other hypotheses.

The first one is simply human psychology. When you think about the nature of a pricing page, it needs to build interest and create curiosity in a way. Even if there’s a transparent pricing page, when users see the prices, they are likely to visit other pages to justify the pricing and understand features better. If there’s no transparent pricing, I don’t think this situation will massively change because in this case, users will probably look at your other pages to understand how much it might cost – as discussed above, for example, if your homepage is full of enterprise logos and you don’t have a PLG motion, then it’s not hard to guess that your product won’t cost just $100. 

My second idea revolves around landing page jails. Landing page jails are landing pages used mostly in PPC campaigns, where the user has no option to navigate other than submitting the demo form; the user doesn’t see the header, footer, or anything else that could take them to another page. The user has only one choice: submitting the form—there’s literally no other option. However, I don’t think this second hypothesis is necessarily a strong one because the use of landing page jails is not that common, hence there’s a really slight chance that it could have impacted the overall data. (Don’t know the origin of this term, but I first heard about it from Gaetano )

My third hypothesis is about the general PPC experience. When it comes to PPC, marketers, including myself, tend to be really conversion-focused and direct their entire PPC traffic to demo pages. This is not necessarily a bad thing, but it also depends on the keyword intent. I see no problem with directing the “HockeyStack competitors” traffic to a demo page where I compare my product with others; but I also know that I only target high-intent keywords. For every account I have audited or managed, I’ve always seen that there’s always this one campaign with low intent but high volume where marketers are too afraid to pause because pausing it certainly will damage their number of MQLs. I don’t and can’t blame them, I understand the pressure they are facing, the pressure to generate more and more leads – but this would also explain the high bounce rates on the demo pages. These low-intent, high-volume keywords direct users to demo landing pages. Since most of them are unqualified, they directly bounce. The not-so-unqualified ones submit the forms, and a small percentage of them actually end up moving through the funnel.

Another interesting data point to look into alongside bounce rate is the average time on page. For the demo pages, the average time on page is 1 minute and 51 seconds, whereas for pricing pages, it’s 1.8x more; meaning that there’s more than a minute difference. This suggests that not only do people who visit pricing pages bounce less, but they also spend more time on the page; regardless of whether pricing is transparent or not.

Let’s summarize what we discussed:

-The pricing pages get 13x more visitors than the demo pages

-The pricing page visitors spend 1.8x more time compared to demo page visitors. 

-Pricing page visitors tend to bounce 80% less. 

On the conversion side though, things change. As mentioned above, pricing pages have an average of a 3.8% conversion rate; for demo pages, this rate jumps to 5.5%. 

Regardless of whether the pricing is transparent or not, demo pages on average have a 1.44x better conversion rate than the pricing pages. 

Apart from the pricing and demo page conversion rates, we found out that the average website conversion rate per unique visitor is 1.1% meaning that for every 91 unique visitors, a B2B SaaS company gets one form submission on average. 

Part III – Do Case Studies Work?

This was initially not going to be in this report, but it was something that we’ve been discussing internally. Companies are spending so much resources to convince their customers for case studies; they create videos, write long articles, and everything, but:

-Do they know if these pages get visitors? 

-Do they follow the metrics of these pages?

-Do they know if these pages actually convert? 

Let’s start with the ratio of website visitors. On average, case studies get less than 1% of all website traffic, to be exact, 0.76%. To compare, pricing pages get 16.5%, and demo pages get 0.91% of the website traffic. This suggests that the traditional website journey funnel where users see the homepage, then product pages, then case studies, then pricing, and then demo is broken (at least for now).

On the bounce rate side; case studies have an average of 53% bounce rate which is slightly better than the average of 59%. But I think we need to consider this from a different perspective. First, let’s think about the definition of bounce rate. “Bounce rate is a metric that measures the percentage of visitors to a website who navigate away from the site after viewing only one page.”  

How does this impact form submissions and pipeline data? 

Case studies, by nature, aim to build trust and should help improve the overall user journey with fewer friction points. If we think about the traditional marketing funnel, case study pages fall under the consideration phase, so the mid-funnel stage. Hence, seeing 1 out of 2 visitors bouncing in the mid-funnel stage is indeed concerning.

What about the people who stick around? This time, we’re seeing a similar pattern to what we saw on the transparent pricing side. On average, case study visitors visit 4.37 pages and spend 3 minutes and 55 seconds. Considering that the average was 2.62 pages and 2 minutes and 35 seconds, this suggests that visitors who don’t bounce engage with the website better.

And how does this impact form submissions and pipeline data?

Once again, we’re doing the lift analysis here to measure the influence of case studies. According to the dataset, website visitors who visit the case study pages without bouncing are actually 8% less likely to submit the demo form, and 22% less likely to submit the pricing form. This is crazy. This data suggests that case study pages actually do more harm than good for most B2B SaaS companies when it comes to form submissions. (This, by any means, doesn’t mean that logos don’t work. Perhaps what visitors need is just to see the customer logos rather than reading through case studies.) 

Things get even more interesting at the pipeline level; although case studies negatively influence the form submissions, it seems like this isn’t the case at the pipeline level. We see that website visitors who read case studies before submitting any forms are 18% more likely to become an opportunity than visitors who didn’t read case studies. This data shows that although case studies don’t have a positive impact for generating form submissions, they actually help increase the opportunity conversion rate. 

Conclusion  

In the beginning of this report, we had three main questions:

– How metrics change when pricing is transparent vs. when it’s not

– What are the average metrics for pricing pages vs. demo pages

– Are case studies really necessary?

It seems like although non-transparent pricing pages have better form submission rates, transparent pricing pages generate more pipeline. As mentioned, I think the main reason for this is that when pricing is hidden, the intent of the submission becomes learning about the pricing without knowing if they can afford it; whereas when the pricing is transparent, the submissions are more qualified since the user already knows if they can afford it – there’s a clear buyer intent to proceed. 

When it comes to demo pages, we’re seeing way better form submission rates than the pricing pages, even though the bounce rate is almost two times higher than the pricing page. Considering the intent levels of these pages, this wasn’t surprising.

For me, the most surprising thing was the state of case study pages and how they actually do more harm than good when it comes to form submissions. But still, this doesn’t mean that these pages aren’t working because we can clearly see that they still influence the pipeline positively.

I think we busted one myth: pricing pages should be transparent. 

And almost-busted another myth: the state of the traditional marketing funnel. It was a widely accepted thought that the funnel was broken, and now we can back this up with more data. Website visitors submit pricing forms, they submit demo forms before checking the case studies—maybe we need to change the way we position the case study pages or the case studies themselves. Maybe this is a sign that we need to test different kinds of pages with different kinds of content. 

I’m going to think about this, and we’ll launch a case study page experiment in the next few weeks. You can subscribe to the Labs Report to learn about the results of this experiment.

I hope you find this report insightful. Please feel free to reach out with any questions or requests for future reports.

If you are our customer and would like to check out your own data, here’s the How-to guide .

Latest Posts

ACV, Sales Cycles, and Sales Reps

The State of First-Touch, Last-Touch, Self-Reported Attribution and Revenue

Uncover What Drives Revenue