August 2, 2016 at 12:53 pm
Eric M Russell (8/2/2016)
Nadrek (8/1/2016)
Eric M Russell (8/1/2016)
2016 can potentially be several orders of magnitude faster than 2008, if you also want to invest in re-architecting your database to leverage things like Clustered ColumnStore, In-Memory OLTP, or Delayed Durability. However, if the argument is that simply running the 2016 upgrade on top of 2008 will yield an easy performance boost, then the argument is less compelling.Enterprise Edition, perhaps; Standard Edition and below, not so much potential gain (no columnstore, no in-memory OLTP, no transparent data/index compression, etc.)
Yes, for a database architect, the only excitement in SQL Server for the past 8 years has been the introduction of new and incrementally improved enterprise features.
What excitement? Enterprise is too expensive for some of us to code to in the first place.
That said, COMPRESS/DECOMRESS is potentially awesome, particularly with views and INSTEAD OF triggers, when you can't get the application changed properly, but the application is one that can work with the INSTEAD OF triggers.
August 2, 2016 at 2:03 pm
My experience so far, 20-30% difference in performance between 2008R2 and 2016 when simply migrating from one to the other.
😎
August 2, 2016 at 2:10 pm
Eirikur Eiriksson (8/2/2016)
My experience so far, 20-30% difference in performance between 2008R2 and 2016 when simply migrating from one to the other.😎
In-place upgrades with no other changes, same hardware, same virtual machine settings, same OS, same storage, same databases, same applications and SQL, etc.?
August 2, 2016 at 2:21 pm
Nadrek (8/2/2016)
Eirikur Eiriksson (8/2/2016)
My experience so far, 20-30% difference in performance between 2008R2 and 2016 when simply migrating from one to the other.😎
In-place upgrades with no other changes, same hardware, same virtual machine settings, same OS, same storage, same databases, same applications and SQL, etc.?
Side by side upgrade on the same hardware and OS (W2012 Srv)
😎
August 2, 2016 at 2:36 pm
Nadrek (8/2/2016)
Eric M Russell (8/2/2016)
Nadrek (8/1/2016)
Eric M Russell (8/1/2016)
2016 can potentially be several orders of magnitude faster than 2008, if you also want to invest in re-architecting your database to leverage things like Clustered ColumnStore, In-Memory OLTP, or Delayed Durability. However, if the argument is that simply running the 2016 upgrade on top of 2008 will yield an easy performance boost, then the argument is less compelling.Enterprise Edition, perhaps; Standard Edition and below, not so much potential gain (no columnstore, no in-memory OLTP, no transparent data/index compression, etc.)
Yes, for a database architect, the only excitement in SQL Server for the past 8 years has been the introduction of new and incrementally improved enterprise features.
What excitement? Enterprise is too expensive for some of us to code to in the first place.
That said, COMPRESS/DECOMRESS is potentially awesome, particularly with views and INSTEAD OF triggers, when you can't get the application changed properly, but the application is one that can work with the INSTEAD OF triggers.
Personally, I've never had use for standard features like geospacial datatypes, filestream, or instead of triggers.
Maybe one day we'll all be able to license enterprise SQL Server features à la carte.
"Do not seek to follow in the footsteps of the wise. Instead, seek what they sought." - Matsuo Basho
August 3, 2016 at 5:43 am
All of the upgrades I have done from one VM to another going from SQL2008/ SQL 2008R2 to 2012 or 2014 I have seen no performance improvement in queries running after upgrades. I have to say though that 90% of these are low end not highly transactional or highly batch running environments.
August 4, 2016 at 7:20 am
From a business perspective, End of Mainstream Support usually is a consideration.
For example, should you replace server(s), and a newer OS is now the standard, you may end up not being able to install the older version.
As features are added and removed in newer versions, you have greater risk that more work will be required to upgrade.
New features and functionality can usually be architected in after.
Sometimes a decremented feature is a larger effort that needs to be addressed beforehand.
28 users, and a manager saying speed is interesting.
Good thing is they appear to be looking for a way to justify getting more current.
So you would probably have some support in setting up a test environment to see first hand if the upgrade would be simple or require a fair amount of work.
Viewing 7 posts - 16 through 21 (of 21 total)
You must be logged in to reply to this topic. Login to reply