Full Moon
Image by Tomasz Frankowski

THE JOURNEY

Our aim is to simplify and lead the way to an efficient and value-adding data center. Therefore, based on our experience, we have developed a model for bringing you, step by step, closer to an automated data center with self-service for your developers. For each step, we have provided links, webinars and other resources to help you on your way. Enjoy!

Objectives of this development phase

  • Understanding how quickly basic knowledge in automation can simplify everyday life for your business

  • To understand the benefits of letting a Source of truth system form the basis for both automation and documentation

  • The possibilities offered via the cloud providers 'and the existing technical platform's various APIs must be investigated

  • The staff should be eager to start implementing workflows and solutions based on what they have learned

Resources

More information

"Containers and Kubernetes sound great, but what about my legacy applications?"
Johan Radivoj is happy to tell you how to proceed to containerize an existing application.
"How can I help our developers when their containers can't reach our on prem systems?"
Rickard Östman has worked a lot with different types of automation solutions, both in the datacenter and in the cloud. He is happy to help with ideas on where you should focus in order to reap results as quickly as possible. Jump on our Slack and ping him there.

Objectives of this development phase

  • To decide which is the preferred automation tool for your organization and what capabilites (and limitations) it offers

  • Senior staff should be able to easily review processes and be able to evaluate whether a task is tricky or easy to automate, i.e. to be able to find the low hanging fruit .

  • Staff should be comfortable retrieving information from available data sources, and be able to use this information as a basis for simpler, automated tasks.

  • Staff should instinctively begin to consider whether there is a simpler and more efficient way of carrying out their everyday tasks.

Resources

More information

Miradot is always available to help taking you further. We offer the Kickstart service to get started quickly, along with a number of other services.

"Both Ansible and Terraform seem great. Which tool is best for our business?"
Sebastian Andersson is happy to tell you which automation tool is best suited for you, depending on what your environment looks like. You find him on ... Exactly, Slack!
"Developers often have to wait for new firewall rules and at every launch there is stress"
Johan Radivoj is happy to talk about how popular technologies such as Github Actions can simplify the rollout of your application while your network can ensure that the right security policy is used and that zero trust can be maintained without compromise.
"How do I get started and what should I focus on?"

Join our slack and talk to Johan. for tips and ideas on how to get started. We know there is a lot to take in.

Objectives of this development phase

  • The staff must fully understand why certain changes can no longer be made manually, as parts of the configuration are now entirely owned by automated systems

  • The staff must work in a more structured way in the event of a change, e.g that a common pipeline must execute an automation tool only after a change has been reviewed by a colleague.

  • Developers must become sufficiently familiar with the infrastructure to know exactly how their application will be deployed in order to make the best use of the infrastructure and monitoring that is in place.

Resources

  • Sebastian shows how a CI-pipeline can allow non-technical staff to make changes to a data center

  • AWS example of how to integrate a Slack channel in a deployment pipeline for peer reviews or similar

  • Examples of how pyATS can be used to ensure that a change in the network was actually implemented in the way it was intended

 

More information

  • Demo on how Kubernetes can interact with ACI

  • Cisco Devnets ACI + Kubernetes sandbox

Help?

Miradot is happy to help with training to make your staff feel secure about the changes that a modernized way of working may entail.

"Sometimes our application crashes in production even though the developers ran their tests"
It is not uncommon for the way developers test their code to need to change when moving into the cloud. Johan Radivoj is happy to tell you how to go about setting up a simple and effective CI / CD pipeline that simplifies the developers' lives while helping them, for example, to comply with the guidelines that apply in your security policy.
"My simple tasks have become very difficult!"
In addition to Miradot's webinar series, there are also our popular courses. Johan Radivoj is a specialist in CI / CD strategies, and he is happy to help your staff identify and overcome any learning thresholds. Everyone should be able to enjoy the luxury of having an environment that works towards the same goals as yourself!
"Our surveillance has not really kept up when things moved out in the cloud"
Ask Sebastian about an introduction to the new monitoring solutions available on the market, or a discussion about how you can best use your existing solution with your new conditions.

Step 2 - Ready for automation

A cloud strategy has been adopted. The IT department has insight into how the rest of the organization works with the cloud, and they use tools such as Terraform and Ansible to manage cloud resources consistently. Communication has been setup between the local data center and the cloud, and relevant parties have agreed on a model for how cloud services and the local data center should coexist. In the local data center, there's an understanding of ​​which existing systems could be managed via APIs. You have started to use a system for DCIM (Datacenter Infrastructure Management) which is perhaps currently mainly used for documentation, but the system has APIs and will therefore be able to support automation tools with so-called Source of truth information.

 

Standardization work has begun and new installations are carried out according to a documented routine. Cloud deliveries take place quickly, but the data center itself is still characterized by slow deliveries and reactive troubleshooting, and therefore continues to slow down business.

Risks

  • Some deliveries are faster as they are now executed with scripts, but as these are typically run locally on staff computers, there is a risk of conflicts, and due to lack of traceability, such situations can pose a threat to operational reliability.

  • Continued risk of human error as most deliveries are still done manually

Degree of maturity

Low. Security has been significantly improved through clear ownership of cloud infrastructure and a standardized approach. There's now a technical foundation and a clear vision for the future development, which will greatly facilitate taking the data center to the next level - A future where staff can focus more on operational benefits than "putting out fires".

Step 3 - A partially automated data center

The organization is now familiar with one or more automation tools and has automated a number of time-consuming and repetitive tasks. The data center is managed in part via CLI / UI and with scripts / playbooks, but now tools for CI / CD are also starting to come into the picture.

 

More efficient changes and faster deliveries are starting to free up time which is now invested in creating business value by allowing staff to learn more about automation and DevOps. A threshold has now been reached where the staff can to a greater extent iterate further over their own processes without the need for external competence. The IT department is to some extent still reactive, but innovation now comes from IT and the developers in a common forum.

 

A seed for a service catalog has been sown, where services such as the Kubernetes cluster are offered both on-prem and in the cloud, and there is a plan for how legacy systems should be life cycled.

 

Traceability and restoring previous configuration are still a problem, but most monitoring and documentation are now part of the automation workflow. Technical staff have found inspiration and started experimenting with new systems and working methods. Some workload has been migrated to the cloud but can communicate with systems on-prem. Overall security has been improved as a result of standardized deliveries strengthening the understanding of how the applications are related, which has led to the start of proper network segmentation.

Risks

  • Parts of the workforce may feel that they are losing control of how the new processes work and are starting to fall behind in development

  • Technical changes can now be implemented at larger scope, as the automated processes enable them to take place at a faster pace than before, and without proper testing this can lead impact availability.

  • The manual administration that still occurs is often related to the tasks that are most difficult to automate, and without clear traceability, such changes can pose a threat to operational reliability

 

Degree of maturity

Average. The benefits of automation are beginning to show for the rest of the business, which is experiencing faster deliveries and fewer incidents.

Step 4 - Fully automated data center

All equipment in the data center can be configured programmatically and a so-called software-defined data center (SDDC) has now been achieved. Your organization is comfortable with its automation strategy and cross-functional groups have been created to further improve existing automation flows/processes, as well as find new ones where needed.

 

Changes are made solely based on structured data from an agreed origin (source of truth), and are executed by CI/CD pipelines. This in combination with changes undergoing an approval process through, for example, a Gitlab merge request, means that you get clear traceability, which greatly facilitates troubleshooting in the event of an incident caused by human error or similar. Now that this level of process standardization has been achieved, it has also become trivial to stage any and all changes in the company's test environment before they are implemented in the production environment.

 

All deliveries are standardized and delivered flawlessly and quickly, both on-prem and in the cloud. Product owners are happy with how fast and easy it is to launch new services. Via monitoring systems, you can get full visibility of how bits and pieces of your infrastructure is communicating, regardless of whether it is a container on-prem or a virtual machine in AWS. The segmentation work has been completed. Workloads can be shared to the cloud to handle peaks in workload or failover scenarios. Time is mostly spent on value-adding tasks such as improving and developing services, establishing feedback loops and listening to what challenges the various business verticals experience in order to be able to address them.

Risks

  • Capacity planning becomes more difficult when consumption becomes easier

Degree of maturity

High. It has an innovative culture with a focus on delivering business value. The data center and the IT department are no longer a bottleneck but rather a driving factor in development. The developers are satisfied and the business sees the IT-department as an enabler.

"The security team has had a hard time maintaining policy since our routing was extended to AWS"
Johan is happy to talk about how to proceed to maintain zero trust in a hybrid cloud environment, and also how to get back on track if you have taken some shortcuts in the past
"The builds take a very long time to run and we encounter bugs quite often despite all the tests"
Do you have a pipeline that you think could work better and run faster with a little more modern technology? Do you also want a second opinion on how you intend to carry out your integration tests? Perfect. Hilight Johan on Slack!

Objectives of this development phase

  • Staff should begin to feel that it is difficult to occupy themselves with delivery-related tasks. Infrastructure maintenance shouldn't require much daily overhaul.

  • The staff should actively collaborate with the company's developers in order to find new value adding services and solutions.

  • Efforts are being made to better package and make both existing and new services more accessible, for example via a portal.

  • Cost calculations are being done based on service and location, which means that the workload can allocated where it works best in terms of performance and cost.

Resources

The company's datacenter infrastructure is now fully automated, and, based on their need for resources, the developers own everything from virtual resources to physical servers in the data center. They know exactly how to go about life cycle management of their applications and how to allocate the resources they need. They can easily get an overview of how their application is doing and how the communication works over the infrastructure's various domains. Documentation and monitoring are included, so that the application owners respectively receives notifications about their services if a problem occurs. If the application is no longer used, the owner will be notified of this, and he or she can return the used resources on their own.

The organization's buyers can monitor the use of resources in both the cloud and its own data center, and can also make forecasts of when new investments need to be made. The IT department can now focus on developing value-adding services for the company's various departments, as well as proactively evaluating the new technologies that the market turns to, in order to be able to continuously improve its own operations.

Risks

  • Boredom within the IT-department

Degree of maturity

Very high.

Step 5 - Self-service

Now what?

We know that all of the above does not apply to you. We know that there is no one-size-fits-all recipe. However, we do know that everyone needs inspiration, guidelines and somewhere to turn if you get stuck. Therefore, we developed this model together with services that in a simple way act as inspiration and provide a predictable way to turn your vision into reality.

KICKSTART

SINGLE TASK

MIGRATION

PROJECT

 
 
 
 

Step 1 - The classic data center

The data center is characterized by technologies such as VLAN's, physical firewalls and virtualization. Most servers are managed individually and there are several systems with diffuse ownership or those with only a single person in the group who knows how they work. Since the development and administration is manual, time-consuming and inefficient, the company's software developers have started looking to the cloud to speed up their deliveries, something that has apparently happened without an overall strategy. Documentation does not always match reality and monitoring for new services and hardware is sometimes overlooked. IT departments are often forced to work reactively and rarely has any time to invest in adding business value.

Risks

  • Competence silos - individuals are business critical

  • The business is slowed down by the IT department

  • Manual configuration is error prone. Will impact service resiliency.

  • Uncontrolled and inadequate security

  • Impatient developers begin to use unsanctioned services, which in turn leads to shadow IT

Degree of maturity

Low. The IT department often becomes a brake pad due to manual work with a lot of “putting out fires”. The business goals are not synchronized with the goals of the IT department, which makes budgeting and decisions difficult.

We challenge ourselves, our customers and the market with new business models, collaborations and solutions. We strive to be the best at what we do and believe in partnerships with others with the same ambition to create a whole. Read more about us

+46 (0)10 - 516 38 10

Follow us on social media!

Miradot AB     |     Borgmästargatan 12, Stockholm, Sweden     |     VAT SE559039952201

© 2020 Miradot AB