Remember Database Mirroring? I do. Vaguely. In particular, I remember working with Database Mirroring (DBM) as part of an interesting -and complex- HADR setup. Visualize a 4-node windows fail-over cluster in Data Center 1 with SQL Server installed on each node (lets call it a 4XActive). We have high-availability covered here via the clustered instance. […]
Au Revoir 2016, Bienvenue 2017! Ca Va! No, I don’t speak French but that never stopped these guys: Foux Da Fa Fa. And this title, as they say in France, has a certain “oh I don’t know what” ring to it. Sacrebleu I was a busy throughout 2016. Mostly working on… well… mostly the work this […]
Say you’re making a big jump and upgrading a database from SQL 2000 to SQL 2014. Of course it only makes sense that this will be a side-by-side upgrade and therefore you’ll need a plan to migrate your DB from the SQL2000 box to the SQL2014 box. SQL2014 won’t allow a restore or attach of […]
Consider a 2-node cluster with 7 instances. Each node has 8CPU (logical) and 64GB of RAM. It’s a (so-called) active\active configuration. Â Two of the instances are especially high-resource-use and are segregated to their preferred node. The other instances are relatively quiet, using minimal CPU and RAM, and are split-up 2 and 3 to their preferred […]
Check this out – the almost last page of a SQL install, Ready to Install. See the configuration file path at the bottom? Recently I have been doing many SQL installs on Azure VMs. As AOAG ‘nodes’, each install needs to be, basically, the same. There are options for imaging Azure VMs but in this case I don’t […]
Take a second to consider this amazing time we live and work in. We are tech pros and we are riding this huge historical wave of the information revolution. This revolution is not even close to over, we are somewhere in the middle of it all and who knows how far it will take us. But […]
Not dead yet Sure, Database Mirroring (DBM) has been surpassed by its younger sibling Availability Groups, its full functionality reserved for Enterprise Edition, and it is marked as a deprecated feature. But it’s not dead yet. It’s still a valuable DR option (especially for Standard Edition) and it is an excellent tool for migration projects […]
Not everything is small in Flatland. Remember that cool old book Flatland? Or maybe you saw the movie? Aside from the mashup title where I’m confounding VLFs and VLDBs this post has really not too much to do with Flatland (sorry). Except that, as you must know by now, I am a square. I’m not the first square to […]
To follow up on my last post, here is a look at how to get the file sizing correct if\when it is determined that multiple tempdb files are a necessity. See that post for a discussion as to why multiple tempdb file might be needed. This post just concerns methods to get there… in this scenario […]
One of the ‘best practice’ practices I like to consider putting in place on a SQL instance is multiple tempdb files to eliminate any chance of tempdb allocation contention. Not that I’ve seen this allocation issue in the wild so much (although Paul Randal mentions it is pretty common in his experience), but having a couple-few, right-sized […]