How has it helped my organization?
All development and pre-production scenarios are now under pay as you go, and are not open all the time, in that way we’re saving a lot of resources, money and we gain fast flexibility to grant new capabilities, computing power, and PoC scenarios.
All of this without compromising production workloads and overall computing power, and any investment.
What is most valuable?
Software designed network capabilities, flexible computing, and managed storage. All of them together make a hybrid datacenter design more flexible for users and IT pros.
SDN capabilities make anyone able to manage and organize a virtualized network in as many levels (VNets and subnets) as you want, securing aces as well. You can organize many VNets and easily interconnect them. You have several and easy ways to connect your virtual DC in Azure with your on-premise DC -- making it easy to have hybrid environments.
Managed storage capabilities, which create a very simple way to create, copy, and replicate local or geo-replicate, it's very simple to assign workloads.
All classic storage configuration settings are now managed by the platform. Huge granularity of compute availability makes it really easy to get appropriate sizing or to change the sizing of actual or future workloads.
What needs improvement?
One of the most important areas for improvement is the administrative part of management. It‘s difficult to manage, all aspects of Azure invoicing, and further pricing vs usage comparison and capacity to move under usage resources to a better positioned.
In this matter, it’s necessary to use third party products or to use good self-management economic tools
What do I think about the stability of the solution?
Sometimes in the management portal, not in the workloads.
What do I think about the scalability of the solution?
Not at the current situation. Not all workloads are working in Azure.
How is customer service and technical support?
In our case, tech support was requested only two times, regarding Azure AD integration issues and special domain resolution issues. It was solved in a good way.
I want to mention a special domain resolution case. It was not easy to solve and was difficult to find a escalation engineer in order to understand the “problem” to fix it.
Which solutions did we use previously?
Because of better integration with AD and Azure AD, Office 365, and IaaS and PaaS Services.
How was the initial setup?
In general, it was straightforward, but was really well analyzed and planned in order to minimize possible problems and complexity.
What's my experience with pricing, setup cost, and licensing?
It’s a good idea to use BYOL if you have an EA. It’s a really noticeable cost reduction.
It’s also interesting to analyze carefully all invoicing costs and workload usages -- to better fit costing scenarios.
Which other solutions did I evaluate?
Yes, we analyzed Amazon AWS and Oracle Cloud.
What other advice do I have?
Be careful, not all workloads are interesting or cost viability to move to Azure as is. In most cases, it will be necessary an important transformation to better fit the Azure ecosystem.
Focus on a first project in order to test all aspects related with platform, providers, own tech capabilities, costs -- that will give you all the tools to decide future plans.
Disclosure: My company has a business relationship with this vendor other than being a customer:
Aug 06 2017