All at sea in cloud migration? These 7 considerations might just save you
1) We moved all our Servers into the cloud, nothing appears to have changed, where’s the benefit?
The cloud won’t innovate on your behalf. If you have no plan for how to modernise your IT estate as part of a transition into cloud, you’re going to miss out on the real key benefits. This is a common mistake companies make when using the lift-and-shift model. There is no doubt this is a quick and valid way to move workloads into public cloud, especially when working with strict and rigid deadlines. An agreed datacentre exit is a good example.
Your end goal should not be to create traditional datacentres in the cloud, why go through all that effort and expense to create something that you already have available to you. The end of a successful lift-and-shift migration should be the catalyst for becoming truly cloud-native and that’s where the real benefit can be unlocked.
2) This is expensive, I don’t understand, I thought cloud would save us money.
A very common misconception is that cloud is simply cheaper in all aspects. This is false. Cloud can be significantly cheaper, however the execution is critical. If you treat your public cloud platform with the same approach you’ve always had to IT infrastructure then cloud becomes costly, and very quickly. Here’s some some useful tips below for the types of changes that can prevent your cloud costs spiralling out of control;
- Identify which workloads can be refactored to be truly cloud native. Traditional IaaS workloads can often be transitioned into PaaS and SaaS offerings that unlock large cost savings. Can websites previously running on virtual machines be reworked to run in App Services? Can Microsoft SQL Databases be migrated into Azure SQL or SQL Managed Instances?
- Use the built-in tools that cloud providers offer to constantly monitor your workloads. Tools such as Azure Advisor provide free recommendations on things such as cost, security and performance to help you better understand your cloud workloads. Virtual Machines running in Azure are often over allocated resources that can be “right-sized” to a different SKU to save costs, for example.
- Use tagging to attribute resources to specific cost-centres, you’ll be much more popular with the finance department if you do! All too often companies have deployed resources into the cloud, costs are starting to grow and they’ve no way of telling which costs are attributed to specific work streams.
3) We don’t need our internal IT staff now, right?
Incorrect. Any migration to public cloud can create concern for IT departments on both sides of the argument. IT leadership can sometimes incorrectly believe a mass culling of IT staff can begin, whilst other members of the team will often believe the cloud is going to render them redundant and no longer required.
The fact is, any successful cloud adoption strategy should place the re-skilling of IT staff at the forefront, and it’s vital this process begins in advance of any migration tasks taking place. Whole IT departments must buy into the process of upskilling and modernising staff to be ready for cloud. Public cloud providers offer free credit to start using their platform, free (and paid for) training events and certification programs to drive this process forward. Mass skill gaps within IT teams is good for nobody.
There is no doubt that more traditional IT roles must evolve as part of the process, age-old tasks such as dealing with Server hardware issues of course will be removed, but that doesn’t mean IT roles as a whole should be.
4) After re-platforming, it turns out our legacy application doesn’t work in the cloud and can’t be refactored to be Cloud native.
This is becoming far more common, companies embark on a journey to migrate legacy applications into the cloud, without fully understanding the implications, and whether it’s technically possible. Not every application can be moved into the cloud, sometimes legacy applications have to remain in situ, it simply unavoidable. Selecting the right workloads to migrate into cloud is vital, a failed migration is costly, needlessly drains the time of technical resources and makes it much more difficult to secure buy-in from leadership teams to move additional applications later down the line.
It’s important to make use of the application assessment toolkits that are available to you. Public cloud providers supply tools such as Azure Migrate to help assess the feasibility of your migration and locate any bumps in the road you may encounter along the way. Often, an estimated costing is provided to help shape your decision on whether migrating is worthwhile and cost effective. One failed migration doesn’t mean that public cloud isn’t for you, it more likely means sufficient pre-planning didn’t take place on your first attempt.
5) Our Cloud migration seems to be taking ages, I thought this was simple.
A common misconception is that cloud is easy, whether migrating workloads or setting up from scratch. The process can’t be that difficult, can it? A cloud migration is no small undertaking, many companies under-estimate the volume of work associated with making a success of their cloud journey.
Whether or not your transition into Azure can be considered a success is not decided by how long the process takes, key stakeholders and interested parties will always want an indication on how long it’s likely to take though. Below is a list of key points for managing the expectation surrounding your migration and how to keep the process moving in a positive direction;
- Split the migration down into small, more manageable work streams. Application by application, or in small groups of applications. It’s very easy to lose track of project status when moving too many workloads at the same time, an initial large upturn in Azure spend on a non-finished project can also raise awkward questions from your finance teams.
- Set realistic goals for your cloud migration, create milestones for the project based on these goals. Using tools such as Boards within Azure DevOps to monitor outstanding tasks and backlogs is a great way to monitor your own progress but also demo key wins and successes to interested parties.
- Use the tools you have available to you to map out a realistic plan. Tools such as Azure migrate should be used to sanity your migration plan, ensuring that any workload you intend to move is suitable before you start.
6) We constantly hear about DevOps and Automation, but I don’t see where it fits for us.
Whenever cloud is mentioned, DevOps and Automation follow not too far behind. Often companies have heard how using automation can streamline processes, save money and remove tedious manual tasks from their day-to-day duties, but they don’t fully understand how it fits into their business.
Automation is a major step forward in modernising your IT estate and becoming truly cloud native.
- Modernise your approach to IT infrastructure, going further than just migrating to cloud native services. Do you need to run your workloads 24 hours a day? Why pay for resources when nobody is working to consume them? Can servers be switched on in the morning and turned off at night? Use Azure Automation accounts to handle this with no human intervention.
- Take Automation one step further by utilising features within Azure DevOps. Do your non-production environments need to available constantly? Make use of CI/CD pipelines, with infrastructure-as-code, to remove whole environments when they are not required and spin them up when they are.
- Embrace DevOps, modernise working practices. Create cross-functional, multi-skilled teams and remove the frustrations caused by the historic Developer vs IT Operations culture.
- Instant access to new regions and markets as they become a requirement. Using repeatable infrastructure-as-code templates and CI/CD pipelines, have new regions online in hours, not weeks/months.
7) Everything works, but is it secure? How do I tell?
One major concern when thinking about cloud is how can they tell the cloud is truly secure? Often, they already have production workloads online, but have no real grasp on whether it is secured correctly or not.
Security is not just about malicious threats from external sources, these are valid concerns, but it also about what guardrails can be put in place to protect against human error and lack of understanding. Cloud platforms can unwittingly allow instant and privileged access to business-critical resources for staff members that shouldn’t have access, if your cloud platform is not governed correctly.
- Create a well governed cloud. Working with the principle of least privilege, use role-based access control to grant users access to exactly what they require, and no more. Custom roles can be created and tailored to suit business needs, these can be granular to the extent that access is granted at specific resource levels, resource groups or subscriptions.
- Familiarise your Security Operations team with the new threats and issues that cloud brings. Understand that cloud comes with a shared responsibility model, there are parts that are fully within your administrative control, but certain services and platform offerings require a greater level of trust in cloud providers to keep you secure.
- Utilise the tools that can highlight security misconfigurations within your cloud platform. Azure Security Centre and Azure Advisor will constantly alert on insecurities and often provide solutions for how these can be resolved. Regularly monitoring and checking this is a very useful way to stay up to date with the status of your cloud security.
- Finally, don’t forget that traditional security best practices still apply. The cloud brings new challenges, but traditional methods for securing your IT estate are still very valid.
So whats is next in your migration?
Get in Touch.
Let’s discuss how we can help with your cloud journey. Our experts are standing by to talk about your migration, modernisation, development and skills challenges.