The short story for this quarter: SQL Server 2019 is on fire, dominating the market.
For the long story: ever wonder how fast people are adopting new versions of SQL Server, or what’s “normal” out there for SQL Server adoption rates? Let’s find out in the summer 2023 version of our SQL ConstantCare® population report.
Out of the thousands of monitored SQL Servers, a whopping 44% are SQL Server 2019! That’s the highest percentage we’ve seen for any version in the 3 years that we’ve been doing this analysis.
I’ve excluded 2008, 2008R2, and the various flavors of Azure SQL DB from that chart because the numbers are all much smaller than SQL Server 2012.
SQL Server 2019 is up 6% from the last quarter, and it’s taking market share from every other version but SQL Server 2022. Every other version’s adoption rate is down 1-2% this quarter.
SQL Server 2017 is now the version that time forgot: folks are just skipping past that version, standardizing their new builds on 2019 rather than 2017. There wasn’t anything wrong with 2017, per se, but it just came out too quickly after 2016. These days, if you’re going to do a new build, I can’t think of a good reason to use 2017. I’ve also updated my Which Version to Use post to reflect that.
After a year of availability, SQL Server 2022’s adoption rate is 4%. That’s somewhat behind SQL Server 2019’s 6% adoption rate of late 2020, but look how big 2019 is today. While I’ve repeatedly gone on record saying 2022 is a mess today, I still think it could be fine long term if two things happen:
- If Microsoft gets the Azure-MI-as-DR thing easy enough for small businesses to use, because companies will adore that, and
- If Microsoft doesn’t release the next version of SQL Server before late 2024 at the earliest. If another version comes out quickly, then 2022 will be the next 2017, skipped along the way. I think this is a safe guess because here in late 2023, we still don’t even have an announcement of the next version, let alone feature lists, public test versions, or a release date. Let’s generously say the next version comes out in 2024 – that still gives SQL Server 2022 several years of time to be the “latest safest version” after vNext ships. And…
- If companies don’t decide to migrate to Azure SQL DB, Managed Instances, or rewrite their applications for Postgres or whatever. It’s conceivable – not likely in the short term, but conceivable – that those competing products might get popular quickly enough that SQL Server 2022 might never catch on. (I’m less worried about this for 2022 than I am for its successor.)
Here’s how adoption is trending over time, with most recent data at the right:
SQL Server 2014 goes out of support in July 2024, so if you’re still running 2014 (or prior) in production, it’s time to make those transition plans to get onto a supported version. 2014 is currently at 7% of the market, and 2012 & prior are at 3%. As those versions continue to taper down, it’ll be interesting to see if SQL Server 2019 grows even further. Wouldn’t it be wild if fully half of the servers out there were running 2019?
These fantastic 2019 adoption numbers really influenced my training class plans for 2024. I’m going to do another round of live classes next year, updating all of my classes to focus on 2019’s features that are widely available. It feels like 2019’s the version companies are standardizing on for the next few years, so it makes sense to standardize our skills on tools that will be available on most of our servers, if not all.
13 Comments. Leave new
The low Azure ratio could also be caused by not so frequent usage of SQL ConstantCare utilities. Imagine when the provider handles most of the maintenance tasks and when the higher HW performance is easier to achieve than any kind of SQL tuning then why to use additional tools?
Brent commented on the low amount of Azure DBs in a previous installment of the report:
https://www.brentozar.com/archive/2022/12/sql-constantcare-population-report-winter-2022/
Quote:
Okay, next up – adoption trends over time. You’re going to be tempted to read something into this chart, but I need to explain something first: we saw a huge drop in Azure SQL DB users for SQL ConstantCare. In the past survey, we had exactly 500 Azure SQL DBs being monitored – and this round, it dropped to just 64. I talked briefly with a couple of the SaaS customers who stopped monitoring their databases, and they both said the same thing: “We’re not going to change the app’s code or indexes based on what you found, so we’re not going to monitor it further.” That’s fair – throwing cloud at it is a perfectly legit strategy.
Unquote:
Great analysis, thank you Brent
Nice data! Thanks for sharing. Here at AWS we see a similar version profile for our customers running SQL Server on EC2:
SQL 2012 | 3%
SQL 2014 | 12%
SQL 2016 | 26%
SQL 2017 | 14%
SQL 2019 | 37%
SQL 2022 | 5%
Interesting, thanks!
Thanks for saying “I’ve excluded 2008, 2008R2, … because the numbers are all much smaller than SQL Server 2012.” Not that I’m confessing a few vendor apps are still run on 2008. Yes they could be moved in compatibility mode but the “fear of breakage” is too great.
These numbers match my shop. 2017 got “middle child” over looked.
SQL 2025, I’m guessing November 2025 at Ignite/SQL PASS, might be a widely adopted.
I just hope for “On prem” solutions. Why? Some business can not shut down nor run remotely (think water treatment plants, 9-1-1, jail systems, etc) FEMA shows up in the wreckage of the city looking for Police/EoC/GIS reports. Municipalities may be an edge case like war ships and subs that need an on prem solution.
While it looks like people are slow to adopt 2022, the graph shows 2019 was similar. In 2020Q3, the adoption rate for 2019 was only 4%, right were 2022 is now in 2023Q3. Both were released in November of their “name year” so those are equivalent time-since-shipped.
Did you read the post? I literally talk about that in the post. You know you can’t just look at the pretty pictures, right?
I did read the post, but.. Hey – pretty, shiny thing!
Look, a squirrel!!!
We are in AWS, and have migrated all of our lower environments and BI servers to SQL 2022/Windows 2022. We are not leveraging the “new” features that have yet to be released and so far, so good. Our current servers are on Windows 2012 R2/SQL 2014, and we are migrating our production servers next month to SQL 2022/Windows 2022, and will follow the advice in “How to Go Live on SQL Server 2022”.
What do you make of the apparent crash of Azure SQL DB from 11% in 2022 Q3 to 2% one quarter later and virtually 0 now? Or am I reading the lower chart wrong?
Brent commented on the low amount of Azure DBs in a previous installment of the report:
https://www.brentozar.com/archive/2022/12/sql-constantcare-population-report-winter-2022/
Thanks, Montro. So others don’t have to look it up, here’s what Brent said then about this in a much longer post:
Okay, next up – adoption trends over time. You’re going to be tempted to read something into this chart, but I need to explain something first: we saw a huge drop in Azure SQL DB users for SQL ConstantCare. In the past survey, we had exactly 500 Azure SQL DBs being monitored – and this round, it dropped to just 64. I talked briefly with a couple of the SaaS customers who stopped monitoring their databases, and they both said the same thing: “We’re not going to change the app’s code or indexes based on what you found, so we’re not going to monitor it further.” That’s fair – throwing cloud at it is a perfectly legit strategy.