Plan a Seamless Transition to Microsoft Fabric
- October 26, 2023
“Data” has undoubtedly been a buzzword. Microsoft recently concluded the preview for Fabric, a comprehensive solution that promises to unify data analytics tools within the Microsoft ecosystem. With Fabric, data scientists, data engineers and data analysts can work together in a common environment from the same data. This means less focus on infrastructure and more focus on data science.
Transitioning to Microsoft Fabric is a strategic move that requires careful planning and execution. Given that Fabric is an expansion of existing Power BI capabilities—by its joining with other Azure AI tools—we should consider the advantages and disadvantages that come with adding such tools to your existing solutions. It can also be helpful to conduct a cost-benefit analysis with the single storage location for the entire toolset and with the reduction in infrastructure decisions.
As a trusted Microsoft partner, Dell Technologies is uniquely positioned to help your organization harness the full potential of Fabric through a comprehensive analysis of your environment to propose strategies based on your unique needs.
If Power BI is your only analytics tool and is meeting all your requirements for data transformation and insight generation, there may be no need to make a change. New technologies require training time and expenses not only for the users of the technologies, but also to determine how those tools fit into your solution architecture. If, however, you have considered how machine learning might provide better predictive or prescriptive analytics or if your data transformations need the big data power Spark provides, it could be time to consider Fabric. Here are some advantages and disadvantages related to a move to the Fabric toolset.
Data transformation options. Power BI alone limits your team to a single transformation tool: Power Query. Fabric expands that to a selection of tools to better fit different scenarios such as Spark, fast copy and data warehouse store procedures and views.
Versioning confusion. Power BI has never supported version control systems such as Git. Power BI does have a deployment pipeline, which continues to work for Fabric-based Power BI objects. Synapse Studio supports using Git repositories to version control the objects in its workspace. This is a loose end that Microsoft has not yet tied—a common versioning and deployment system for all the components of Fabric. If your team is segmented by roles, each team member can focus on their versioning. If, however, you have a smaller team with multiple hats being worn by each team member, there will be some learning to keep versioning straight.
A common storage location and format for all tools, named “OneLake” within Fabric, gives data scientists, data engineers and analysts the opportunity to all work on and with the same data. The advantages include reduced data duplication and a common location for data set endorsement through certification or promotion. The pitfalls include the need to carefully plan security and architectures that support dev/test/prod environments for your various roles.
Much of the documentation around Fabric labels it or parts of it as a SaaS environment. Many of the Azure Synapse configurations are simplified or handled behind the scenes for users of Fabric. Previously you would have to decide different options for the various tools under the Azure Synapse umbrella (Data Warehouse Units, Spark Pools, etc.). Fabric, instead, asks customers only to select the compute and storage that support all tools in the Fabric capacity. You still must monitor your compute and storage usage and adjust your chosen environment based on actual usage, but the decisions are reduced and thus simplified.
For some scenarios, separate billing and compute may be preferred over the combined compute of Fabric. For example, if your solution heavily depends on Data Warehouse and has no Spark pools, the separate compute selections could be more performant for comparable or less cost.
For the foreseeable future, these separate billing and configuration options are still available, so teams can decide if they want to fine tune different infrastructure elements or if they want the combined toolset that Fabric provides.
Having been a close partner of Microsoft for over 35 years, Dell has deep design, implementation and troubleshooting knowledge and skills across Microsoft’s product line. Dell Technologies consulting experts can guide you to the right solution that meets your specific needs.
Dell can help your organization make the most of its move to Fabric from a variety of angles including training, advisory workshops, staff augmentation, health checks and complete solution design and development.
If your team needs to try learning and prototyping with Fabric now, but you do not want to incur the costs, you can use the 60-day free trial. Be sure to schedule time for your team to use the new tools and features. Prior to sign-up for the 60 days, you may even consider using a partner to help with your learning curve and prototyping. Coordinating the start of a partner relationship with your trial of Fabric could aid you to get the most from the trial period. Reach out to your Dell Services Sales Executive today to discover how Dell Consulting Services can help you get started.