Ever wonder how fast people are adopting new versions of SQL Server, or what’s “normal” out there for SQL Server adoption rates, hardware sizes, or numbers of databases? Let’s find out in the summer 2021 version of our SQL ConstantCare® population report.
Out of the 3,808 servers sending in data recently, the most popular version of SQL Server is still 2016:
When you group versions together it makes for interesting metrics:
- Now that SQL Server 2016 mainstream support is over today, only 34% of installs are under mainstream support with 2017 & 2019.
- The combination of 2016, 2017, and 2019 make up 67% of the market, but…
- The combination of 2014, 2016, and 2017 make up 74% of market share right now, exactly the same as 6 months ago! That’s kinda wild.
SQL Server 2008 and 2008R2’s market share stayed about the same, but 2012 went down by about 1/4, and 2014 down by about 1/10th. 2016 and 2017 stayed stable.
The newer products are at the top of this chart, and the new data’s at the right, so you can see the new stuff gradually pushing down the old stuff over time:
SQL Server 2019’s adoption has almost doubled over the last 6 months. In the last report, I predicted that come mid-2021, 2019 (currently 11%) would have more market share than SQL Server 2012 (currently 9%), and that turned out to be true.
Azure SQL DB’s share has actually gone way down, not up, but I don’t think that’s an indication of the wider market. I’ve had a few Software-as-a-Service (SaaS) companies try SQL ConstantCare® and bail because they get too many recommendations, since we send them for each connection. There’s probably a small market for a SaaS-focused monitoring tool that consolidates database metrics across all your databases and gives centralized recommendations about overall index usage and query workloads, and I’ll blog about that at some point. SQL ConstantCare® definitely isn’t that tool, and I don’t intend to build that tool, either.
SQL Server 2016 is still the juggernaut, with 1 in 3 instances overall.
It looks like cloud adoption is very poor at the moment.
I wonder if there will be any incentives from MS to move to cloud faster?
You can’t gauge cloud adoption from this – some of these SQL Server instances are deployed as VMs in the cloud, like Azure VMs and Amazon EC2.
Azure managed instance is hell a lot costly as compared to Azure SQL DB. No wonder its stuck at 0%. At least my client was rich so we use it for our critical app. After that gigantic paywall its a great feature imo
Costlier & slower, unless you want it to be even costlier.
Why is 2016 so steady? Is it the Windows XP of SQL server versions?
Companies are still doing a lot of new 2016 deployments.
1 out of 3 is still 2014 or older? Cool!
2016 seems fairly stable, has most of the features we want and the ball got rolling to adapt it years ago. So, there is little need to move to 2019…unless of course you want MS support. :/
I’ve seen more than one SQL Server 2019 with their compatibility level still suck to 110. Companies say they have upgraded but they still use the old algorithm.
That’s totally okay, and I talk about why in my Mastering Query Tuning class. I’d even argue that it’s a good default starting point, and I show why with demos. See you in class!
I think it would be really interesting to see how many of the newer sql instances are on windows and how many are on linux. That aspect might be slowing down the transitioning
Out of the entire population, only 3 of the servers run Linux.
Is anyone still using anything older than 2008? I don’t know if you’ve excluded it because there’s no data or whether like a bad past relationship you’d just like to pretend it never happened.
(No, I don’t have any 2005 still, just curious!)
SQL ConstantCare isn’t compatible with pre-2008 servers.
Just wondering if you’ll be releasing another report of this nature in 2022?