If you run an IT group, are an admin, or a SQL Server DBA – it seems change is hitting you from every direction at once. With SQL Server 2008 EOS around the corner, here’s what you need to know about to get ahead of the game.
What a time to be in IT! You’re facing a change in server architecture and deployment with hyperconverged infrastructure; a change in storage architecture with software-defined storage, and another wave of SQL Server migration and consolidation with the imminent SQL Server 2008 EOS (End of Support). All at the same time!
While the cloud is all the rage, even the pundits have stopped saying “move everything to the cloud” – the cloud providers and their customers realize most companies will continue running some workloads on-premises in their own data centers, for the foreseeable future.
Yet did that acknowledgement trigger a big budget windfall for your IT team and data center? Not likely; you’re still expected to do more with less.
That’s where two of the big changes come in: hyperconverged infrastructure and software-defined storage. One of the big reasons they’ve become powerful industry trends is because they can save your company money.
The Rise of Hyperconverged Infrastructure (HCI)
Hyperconverged infrastructure or HCI simplifies purchasing and deployment by using one consistent “commodity” server (sold in high volumes, not specialized), in a consistent configuration (which you can define), that you can buy and deploy repeatedly. That consistency should drive up utilization – when you need more capacity, you buy another server that’s just like the others.
HCI typically runs in a clustered configuration with workloads spread across all nodes of the cluster, so you should get increased reliability, and can get some basic high availability as side effects.
HCI and software-defined storage go hand-in-hand – typically each node of the HCI cluster has the same storage configuration. (That is, each server in the cluster has the same type and capacity of HDDs and SSDs.) All the storage across the cluster is logically combined (“pooled”) and managed by software.
Software Defined Storage with Storage Spaces Direct (S2D)
Storage Spaces Direct is the software-defined storage technology Microsoft includes in Windows Server 2016 (Datacenter Edition only).
The WSSD Solutions catalog lists dozens of validated configurations available from familiar OEMs and ODMs that are ready to plug in and run. (While a DIY S2D environment is possible and allowed, 90% utilization in production is common so getting the configuration just right is important, and challenging.)
Software-defined storage is not about mimicking a traditional SAN, although admins often include the SAN feature list as a purchasing criteria, and fortunately many features you’d expect from a SAN are supported by SDS solutions like S2D. Microsoft has information on S2D at aka.ms/s2d that I won’t try to duplicate.
Where S2D can beat a traditional SAN is low cost, flexibility and ease of management. For example, Microsoft recently introduced Windows Admin Center, a new Windows Server management tool focused on making it easy for server admin to manage the HCI environment, including S2D. No specialized SAN management skills required!
To summarize, so far we have money-saving hyperconverged infrastructure, money-saving Storage Spaces Direct, and easy-to-manage storage. Where does SQL Server 2008 EOS fit into this story?
SQL Server 2008 EOS, Meet HCI and S2D
You may remember when SQL Server 2005 reached the End Of Support (EOS) in April 2016; millions of SQL Server workloads were migrated, many to Windows Server 2008 or 2008 R2. Now it’s SQL Server 2008/R2 that will reach its EOS on July 9, 2019, which will trigger millions of SQL Server workloads migrating to the current SQL Server release, SQL Server 2017.
A key concern: the cost of software licenses and servers.
If you can get fewer cores to do more work, you’d need fewer cores to run your workloads, and fewer per-core licenses. You could save a lot of money.
How can you do this with S2D? In my next blogs I will show you how to run your migrated SQL Server 2008 workloads on a Storage Spaces Direct cluster in the most optimal way. I will look at best practices of how to configure the caching tier and capacity tier optimally for SQL Server OLTP consolidation.
Make sure to subscribe below to be notified of new posts published. Here are more resources to help the outcome of your SQL Server 2008 EOS be a successful one: