Data sources are certainly something that should be on nearly everyone’s radar. With more devices, more analysis and more utility coming from information than ever before, it’s pretty obvious that our pendulum associated with being a data professional will have yet more swings in the direction of managing performance as things expand.
This post from was interesting as they tried to point out the requirements for SQL Server (specifically) to be able to handle these data loads. I thought the headline was a bit of click-bait (obviously it worked; I clicked) but the article solved it’s own problems, talking about scale in processing, scale in storage and so-on having many options.
But I think the points made are very important planning points. Specifically working hard to get ahead of the curve of requirements, so your systems can provide that storage and functionality.
I’ve personally learned the hard way about a few different limitations and differences between cloud providers when it comes to these very things.
For example, scaling up storage on AWS with a SQL Server instance is an issue; it’s not a pretty process, so they coach you to make sure you have the storage you anticipate you will need, when you set up the instance. This is less than ideal, and I’m sure it’ll be solved and corrected, but it sets aside a couple of the benefits of cloud storage – get what you need, expand forever. Neither of these is quite true yet with SQL Server storage.
With processing power, you can easily scale up and down the instance, and this is true of Azure and AWS type installations. This is what is a very powerful capability and helps answer the call for all of the data sources… and uses.
It’s important to take into consideration how you’ll be using things, then asking lots of implementation questions as you choose your weapons. There are differences, and things are moving quickly. I think this is what’s (legitimately) driving a lot of the multi-vendor cloud solutions. There are pieces of each that are outstanding for specific applications and environments.
Now if we can just get some tools that help us manage in a multi-cloud-provider way.