Running Database Workloads on Azure: The SQL Server Migration Opportunity for Hosting Partners
Greetings Microsoft Hosting Partners – and Happy Holidays! As we look forward to 2016, I wanted to share some thoughts with you on the many exciting opportunities for you to present new services and offerings to our mutual customers in the coming year.
In our last blog entry, we discussed SQL Server 2005 end-of-life and how migrating customers to a modernized data platform with Microsoft SQL Server 2014 in your own datacenters is a great opportunity to bring customers into cloud-hosted workloads, whether it’s on your own private cloud or in Microsoft’s public cloud, Azure – and Partner Technology Strategist Sarah Barela shared some great best practices.
Today I’d like to talk about the benefits of using Microsoft Azure rather than your own infrastructure to migrate customer legacy workloads to Server 2014 and again share some insights from Sarah. Here’s what she had to say:
SQL Server workloads can be delivered in Azure Virtual Machines (VMs):
- As preconfigured, pre-patched Windows Server/SQL Server 2014 instances
- Start with D and DS-Class VMs for SQL VMs in Azure
- A specific VM configuration is offered for SQL Server OLTP scenarios
- Use the G-Class VM for extremely large workloads
- Save on infrastructure and Azure costs by being able to shut down servers during off-hours using automation technology
SQL Server workloads can also be delivered in the Azure SQL Service:
- Using a limited set of SQL 2014 code
- HA is already built into the platform, so you don't need availability groups or clustering
- Azure SQL Service should be considered as a database for new applications
- No updates to underlying data platform are needed over time, as you are always running the latest version of the platform with the Azure SQL Service
For more information on SQL Server performance best practices with Azure Virtual Machines, make sure to check out this helpful document on the Azure website.
[View:~/cfs-file.ashx/__key/communityserver-blogs-components-weblogfiles/00-00-00-81-23/SQL_2D00_2005_2D00_EOL_2D00_Podcast_2D00_2.mp3:550:0]