October 21, 2019 at 7:55 pm
Hello Folks,
We are planning to virtualize for one of our SQL environments. As part of this project, we are planning to upgrade SQL as well. Current version/Ed is SQL 2014 Enterprise.
Do you recommend to upgrade to SQL 2016 or SQL 2017 or SQL 2019?
Please share your valuable thoughts... Thanks!
October 22, 2019 at 8:59 am
Hi there,
My stance is always upgrade to the latest version if you can. Maybe that's not for everyone but with enough testing there shouldn't be any problems. Also make sure you check out all the best practices for visualizing a SQL server. What virtual environment are you planning to use?
I'm currently building a lab to plan our next migration from 2016 to 2019. Migrating AGs is proving interesting.
Andrew
October 22, 2019 at 9:05 am
I'd go for 2017, personally. Don't bother with 2016 in that scenario.
AndrewL65, do you mind me asking what you're finding interesting about it? (Also, is it true that the cardinality estimator has changed again, or am I imagining it?)
October 22, 2019 at 9:09 am
SQL 2019 is in very early days
I would suggest SQL 2017 as its the latest stable version available.
SQL 2019 has, in addition, some integration features with external sources but it's in early stages still.
I'd aim for 2017. You can then add different versions of the SQL server to the AG and upgrades will be easier. Clusters also optional which is interesting but I'd probably still cluster. I'd look up current issues with each version and see if there are any show stoppers. 2019 is not released yet but almost certainly going to be announced at PASS.
SQL Database Recovery Expert 🙂
October 22, 2019 at 9:28 am
I've not got as far into it yet as looking at the cardinality estimator but main thing that stuck out for me was online functions being considered as part of an execution plan. Had issues with the community preview version because of that and it was faulting on some of our code I tried on it. Seems stable enough now its release candidate 1 though.
My current issue if deciding if I should use a distributed AG when migrating. Still not sure what best practice if regarding that. Has anyone done a 2016 to 2017 migration with AGs?
Our upgrade will be next year so still plenty of time for testing
October 22, 2019 at 12:35 pm
I'm right now migrating SQL Server from 2012 / 2014 to 2017 and I wish I could go for 2019 right away (I'd love to have all the csv files transfered to Hadoop right away instead of some external service managing files for DWH Import) but I might have to do that later as all that's open right now is testing by the business. Wouldn't bother with 2016 either, will just force you to upgrade earlier again.
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply