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 2020 version of our SQL ConstantCare® population report.
The most popular version of SQL Server is 2016. The combination of 2014, 2016, and 2017 make up 75% of market share right now:
But companies are gradually adopting newer versions and replacing their older servers. Here’s how market share changed over the last 6 months:
This quarter, shops continued to gradually replace their older versions:
- SQL Server 2019: 4% of the market, up from 1% 6 months ago
- SQL Server 2017: 24% of the market, up from 18% from 6 months ago
- SQL Server 2016: fairly static at 34%
- SQL Server 2014, 2012, and 2008R2 all declined at the expense of 2017 & 2019
- Azure SQL DB and Managed Instances didn’t make any headway, still at 1% of the total population
- SQL Server 2016 is still the juggernaut, with 1 in 3 instances overall.
- SQL Server 2019 still has less market adoption than SQL Server 2008R2.
So why is Azure SQL DB making so little headway here?
In this data, Azure SQL DB & MI still have less market adoption than SQL Server 2008, a 12-year-old and wildly unsupported version, let alone 2008R2.
I’ll be the first to point out that there’s a natural bias in the source of the data. Our SQL ConstantCare® service lets users send data in about their databases and get health & performance advice via email. SQL ConstantCare® isn’t a full picture of the SQL Server ecosystem: it’s marketed to you, my readers, and specifically the subset of readers who want an inexpensive second opinion about their database performance & health. Conceivably, people running in Azure SQL DB don’t need a second opinion about their database’s health: the patching, uptime, and backups are all handled by Microsoft, and you can’t get much better than that.
However, Azure SQL DB performance – just like any database’s performance – is still a real problem, and I know that from the consulting side of the business. You can scale a long way just by throwing money at it, but eventually your wallet starts to buckle, and you need to fix the queries and/or the indexes.
If I’m wrong – and I certainly could be, I just don’t know – Microsoft could fix this perception by publishing adoption data. SQL Server phones home by default, so they could easily publish adoption rates for different versions. (If they just published download numbers, that wouldn’t really mean anything – lots of folks download stuff they never use.) The fact that they don’t publish metrics makes me suspect that my report’s numbers do give a rough idea of the wider population: 2016 is the big behemoth right now, and the adoption rates for 2019 and Azure SQL DB just don’t hold a candle to it.
I still think platform-as-a-service databases like Azure SQL DB and AWS Aurora are the future king of the market – but just emphasis on future. And at this adoption rate, the future is still at least 5-10 years away.
I’ll ask you, dear reader: why aren’t you migrating to Azure SQL DB?