Some organizations do not want to store their data on infrastructure they don't fully control. Local business data deployment is a choice for them. Local business data deployment is a deployment model where the key components of Dynamics 365 for Finance and Operations, Enterprise edition are deployed on the customer's local data center. This model is suitable for customers who are not ready for the cloud journey due to regulatory reasons or their existing data center investments.

As explained earlier in this chapter, local business data deployment still uses LCS for deployment and management of Dynamics 365 for Finance and Operations, Enterprise edition. However, there is no business data processing outside the customer's or their partner's data center. In this model, the customer/partner is responsible for managing the infrastructure and disaster recovery. Power BI or cloud intelligence is optional for local business data, as these options need business data in the cloud. Customers will have an option to enable it by allowing the business data to be replicated in the cloud. The following diagram represents a local business data deployment model:

As shown in the diagram, local business data means that the finance and operations application deployment is limited to the customer data center. LCS in the cloud is still used for deployment and management, where the configuration, customization, and telemetry data is synchronized to the cloud to enable management and diagnostics.

The following are the key highlights of local business data deployment:

• Disconnected data centers
• Local data residency
• Capitalize HW investments
• Single instance only

The following are the key components that enable you to host your solution on the premises:

• Azure Service Fabric:
  • Microsoft Azure Service Fabric standalone clusters deployed on Windows 2016 server
  • SMB 3.0 cluster (Server Message Block)
• Azure Active Directory (AAD):
  • Authentication for LCS
  • Authentication for VSTS
• OS:
  • Windows 2016 server (standard or data center)
  • Virtual machines running on Hyper-V
• SQL:
  • 64 bit only version
  • SQL 2016 Standard/Enterprise
  • Always ON cluster
  • Bandwidth: minimum 100 MBps with AOS
  • SAN (storage area network optimized for OLTP throughputs)
  • SAN IOPS: minimum 2,000 input/output operations per second (IOPS)
  • Cores: recommended minimum 8 cores for production in dedicated mode
  • RAM: recommended minimum 24 GB for production
  • Hard disk should be solid-state drive (SSD)
  • Virtualization
• Local windows server active directory:
  • Active Directory Domain Services (AD DS)
  • Active Directory Federation Services (AD FS) 3.0
• LCS:
  • Application management
  • Machine deployment
  • Code deployment
  • Servicing operations
  • Environment orchestrator
  • Primary fabric service requiring at least three VMs
• AOS:
  • Should be co-located with SQL
  • Bandwidth: minimum 50 KBps per user
  • Latency:
    • minimum < 250–300ms for clients
    • < 1ms in LAN only with SQL
  • Cores: standard box – 4 to 24 cores
  • RAM: minimum 16 GB per AOS instance
  • Virtualization
  • Typical AOS-to-SQL core ratio 3:1 for the primary SQL Server
• Others
  • Development experience continues to be the same as in cloud deployments through 1-box VHDs, which may need Microsoft Remote Desktop and Visual Studio licenses
  • MR/SSRS (local)
  • Web client (able to access intranet IIS websites)

Let's now compare the deployment options that will help you choose the best fit for your business.