Checkmark=facebooklinklinkedin+Right-facing arrowCheckmarkDown-pointing chevronLeft-pointing chevronRight-pointing chevrontwitter

4 Reasons why to use Docker

Containerization is one of the fastest growing technologies in the software world. One of the leading companies that provide a software container platform is Docker. But why is containerization better for you and what makes Docker so suitable for the Nerdalize Cloud?

At Nerdalize, our biggest challenge is to distribute all the computations from our customers over our CloudBoxes which are scattered over the Netherlands. The portability of containers makes it possible to run jobs on multiple CloudBoxes without compatibility issues.

By utilizing Docker’s technology, we have created a platform and ecosystem which allows us to scale up your computations. You simply provide the Docker image and we provide the computing resources.

Docker logo

The use of Docker isn’t only important for us to distribute compute power, it’s also important for you as a cloud customer. Currently, the cloud market is expensive and rather non-transparent. We believe it’s important that the cloud market becomes more transparent and flexible, which can be realized by using containers to run your computations. Below you’ll find four ways in which Docker containers help you:

1. Portability avoids vendor lock-in at one cloud provider

As I said before the portability of Docker is important to set up our infrastructure, but it’s also important for you as a cloud customer. The Docker image itself contains everything it needs in order to run a specific process. It’s independent of its environment, which gives you a lot of flexibility in choosing the environment you want to run it in. This makes it reliable to run processes in various environments at different cloud providers. And most importantly it’ll be easy to retrieve your data afterwards, avoiding vendor lock-in at one cloud provider and saving you much time and effort.

2. Transparency of costs and performance

Docker containers are based on open standards and it’s supported by most providers. This enables you to run the same container at different cloud providers and get feedback on the performance of different machines. The open standards for containers make it easy for you to benchmark several providers, which makes the costs and performance transparent.

We can also help you to benchmark your process. Our benchmarking tool will help you choose the instance with the best price/performance for your compute job. You’ll receive an overview of the price/performance and runtime per cloud provider for your specific project. This makes it easier for you to choose a cloud provider.

Do you want to know more about our benchmarking tool? Read our blog post on how to budget cloud expenses.

3. Use resources efficiently for multiple processes to save time and reduce cloud expenses

The memory, CPU and disk usage of containers can be accurately controlled. Also, they don’t need a separate operating system, so multiple containers can run simultaneously. Both reducing the total amount of hardware that is used.

Another advantage of containers is that it’s possible to run multiple processes in one container. You can set up a dependency order in the container, which enables you to run a certain simulation and let it directly continue post-processing with a different software package. You don’t need to install any of the software packages since they are already in the container. That makes it much easier to send your complete process to the cloud. Saving you time and effort, because the installation of multiple packages can be skipped.

The ease of controlling the resource usage and the fact that containers run few overhead processes, enables very efficient hardware use. Thereby, we can offer you very affordable pricing for cloud computing. Also, more efficient use of hardware makes computations more sustainable, because less energy is used for the same computation.

4. Easy to maintain and share different versions of software

At Nerdalize we want to make computing easy and straightforward. Because many developers use Docker, new Docker images become available constantly. Many software vendors publish their own Docker containers or supply an installation guide for you to make your own images. This means that you can easily adjust and use existing Docker images for your computations.

Additionally, it’s important that you can easily maintain your containers during a project. It can be essential to quickly fix a bug, add a new feature during the process or apply a new version of an application. Containers can often be run with just a single command. This makes it easy to share your work with your colleagues by just sharing the Docker container itself. You can write your code locally and share the image with your colleagues via Docker Hub or other image repositories. As a result, you can easily collaborate without any version or compatibility issues. A great way to improve productivity when working in teams!

Important to keep in mind

Using Docker enables us to make the cloud market transparent and flexible. It’s a great technology for cloud computing, but it’s relatively new and therefore constantly developing. For some software applications, such as Windows software or applications with complex licensing structures, it can still be hard to create a Docker image. However, when your software application is suitable, Docker has many benefits. Giving you enough reasons to start looking into the technology!

Do you need help to set up a Docker image? Find out how we can support you. We offer in-house training sessions to organizations as well as Docker workshops and we’re always happy to answer your questions.

About our cloud
Liesanne Wieleman
posted this August 30

How do you even budget cloud expenses?

This cost comparison between cloud providers is not easy to make. At Nerdalize, we find it important to help our customers make the right consideration between different providers and instance types.

To make this decision easier for you, we have created a benchmark tool. This tool will give you an overview of the runtime and price/performance for different cloud providers, which makes it possible for you to choose the most suitable cloud provider for your compute job.

Do you want to use the cloud, but you do not know where to start?

Currently, there are a lot of cloud providers with different instances, prices and additional services. There are variable and fixed factors that could influence your choice of cloud provider. Some of these variable factors are price, performance and runtime.

Benchmark badges

The performance of a server is influenced by several factors, which include the number and type of CPUs, storage and RAM. Each compute job has a different specification and thereby needs a different configuration. Some compute jobs might need a lot of RAM, where other jobs are more CPU intensive. Additionally, not all compute jobs necessarily perform better on more CPUs. Therefore it is important to choose an instance that matches the specifications of your job. By choosing the right infrastructure for your job, it is possible to increase the performance and thereby decrease the runtime in order to lower your costs.

Benchmark Report: Plan your own project now!

At Nerdalize we developed a benchmarking tool that runs a small sample of your compute project with a specific software program and sample dataset on different instances at different cloud providers. Comparing the performance of this sample job at different providers enables us to give you an overview of the price/performance per cloud provider for your specific project.

Cost per job
providers/awsproviders/azureproviders/gceproviders/nerdalize

As a result, you get a report with the runtime and price per instance for your sample job. Whether you are looking for the fastest machine or the best price/performance, this overview enables you to choose the right instance and provider for your project. The benchmark report makes it possible to compare the performance and costs of your compute job on multiple cloud providers, such as the Google Compute Engine, AWS, Azure and Nerdalize.

The benchmark report shows you the variables runtime and cost per job, so you can choose the optimal instance for you. Move the slider between runtime and costs and select the optimal instance for you, just like Erik did:

This is really wonderful! First, the Nerdalize Benchmark tool is really clear and well designed… fun to play with. Second, the results are super promising.

Erik Schultes

The benchmark report does not only help you to compare several cloud providers, it is also a great tool to help you forecast the total runtime and cost of your specific project. This helps you plan the total runtime and budget of your compute project and communicate it to your customers or colleagues.

It is important to carefully think about your choice in cloud provider and take into account all the factors. Are you not sure which cloud provider and what factors suit your compute job? Get a benchmark report or get in touch! And find out which cloud provider is the most suitable for your project.

About our cloud
Liesanne Wieleman
posted this July 19

Cloud made easy for hydro engineers

Last week Nerdalize got the opportunity to present our cloud platform to Deltares users. Twice a year, Deltares organizes the Deltares Software Days, an event to meet up with users of Deltares software and talk about the latest developments and trends in the field of hydrology. For the past 3 DSDs we were invited to present our cloud platform. These are always very interesting moments to learn more about our users and their cloud needs!

Together with users of Deltares software we developed our platform in a way that makes it very easy for engineers to run their hydro simulations on our cloud resources. Like in other industry segments, you can see an increasing need for computing power in hydro engineering. With needs for higher resolutions and bigger simulations, projects often demand more computing power than an internal cluster can offer in order to meet deadlines.

In close collaboration with hydro engineers we developed a platform:
  • that decreases the time needed to run hydro models in the cloud
  • that offers computing power with the best price performance
  • without vendor lock-in at a specific cloud provider

Deltares User Research
Using Deltares software in the cloud

This DSD we showed users our new Command Line Interface. We developed a workflow for hydro engineers in order to use Deltares software on our cloud resources without needing support from the IT department. This enables hydro engineers to scale up to the cloud and lower the runtime of their projects without having to spend time and effort in setting up the environment. And use compute power of the most affordable cloud provider out there!

About our cloudcustomers
Maaike Stoops
posted this June 21