QuoStar’s Rob Rutherford shares a few helpful hints and tips.
Certain issues can arise around cloud adoption. However the risks can be mitigated when you know what to look out for.
The increasing popularity of cloud services and software.
There’s been a huge move onto the cloud recently, particularly around providers such as Microsoft’s Azure and the Modern Workplace stack. Microsoft Azure has, for example, reported growth of 51% in Q4 of 2021. This is a huge leap as businesses look to the cloud to improve their hybrid working environments and security.
Many organisations already on the cloud have been transitioning into a public-private hybrid model, or between two public providers to get the right workloads onto the platform to balance performance, security, and cost.
Inter-cloud high-availability and Disaster Recovery environments are areas where we’ve also seen a lot of interest, in order to protect cloud platforms from the failure of one provider or environment. In the longer term – considering state-sponsored attacks and skirmishes will likely become more prevalent in the future – we can expect to see businesses relying more and more on those types of inter-related environments to ensure a greater level of protection.
Cloud adoption: the challenges and barriers.
IT used to be solely seen as a supporting element of the business. One of the main challenges to cloud adoption now lies in the fact that many organisations have built up a technical debt over the last five years. They are trying to transform their businesses rapidly to catch up with the digitalisation of their sector.
Without a true grasp on how to deliver transformation strategies, many businesses have not been undertaking business focused requirements, analysis or mapping projects into a clear roadmap. Instead, they’ve tried working things out themselves. Often this can hold organisations back as they struggle with interoperability and performance issues.
A significant challenge in the actual process of migrating to the cloud is finding a reputable and experienced partner to assist in the cloud transition journey. Unfortunately, on the one side, a lot of cloud providers simply focus on getting the deal signed, and not necessarily on delivery. On the other side, many buyers are too focused on flat costs. Buyers may end up choosing those providers who appear to be the cheapest on paper. As a consultancy, we’re then often brought in to unpick a situation that has been created by rushing through deals.
Balance between cost efficiency and performance.
Preparation is key to any migration. It’s essential that an organisation doesn’t simply take the word of a salesperson on how long a migration is going to take, how much it’s going to cost in the first instance, and then on an ongoing basis.
All too often migration projects need to be pulled back on track. Performance issues might need to be addressed in an environment (ideally without a price increase from a customer’s point of view). Which can be hard – if not impossible. Quite frequently the ROI stated in a cloud provider’s proposal falls away as the realities of a complex workload bloom. This is where an organisation has signed an order and gone through significant (and often horrendous) migration, only to be left with a screaming user base and/or customers. The pain they’ve incurred often means they make drastic decisions. Such as wiping out ROIs, increasing the security risk profile or getting into further contractual obligations.
Most cloud providers will give an organisation some form of free trial on a workload. The larger the workload the more complex this gets; however commercial deals can be made.
It’s critical that organisations do their due diligence and build in contractual obligations. This will help to ensure the supplier delivers the desired outcome they expect. It may even be advisable to bring in external consultants, and/or lawyers to take some accountability for the project delivery. Especially when looking at large scale migrations.
Avoiding performance issues during cloud migration.
It’s always a good idea to over-resource when undertaking initial migrations. A large percentage of environments take up more resource, especially in the early stages of a heavy migration.
One of the biggest areas people typically under-resource is disk speed, in terms of IOPS (Input/Output operations Per Second). Too many organisations throw memory (RAM) and processor power at an underperforming environment. While it’s the disk speed that is the bottle neck. This isn’t a new cloud-related issue. However, too many IT teams spend time chasing their tales when the speed issues are in fact related to disk IO. Often many cloud provider support teams don’t seem to understand this in the lower support tiers, so be aware.
It’s worth being careful if you are paying for disk IO and/or network ingress or egress traffic. This is often where cloud costs start to spiral away from the original quotes first agreed upon. The public cloud often appears cheaper than private platforms when you go light on these costs. It’s worth checking these beforehand. It’s important testing your environment under load or having the cloud supplier make some guarantees around costs.