A great invitation this month from Deborah Melkin. I thought this was very apt, as I use this phrase often. Many of us do.
This was a tougher post to write because, well, too many choices. However, I sat down to try and give an example of where this comes into play for me.
Everyone should write a post, or a few posts, on this topic. Your decision making and reasoning are important things to share with others.
The Devil is in the Details
I deal with quite a few proof of concepts and questions about altering database development practices. One of the questions I’m often asked is how does a particular group get started with automating their database deployments.
That’s really the answer, and I have to start to ask them questions about their pains. What is the driver for change? What things don’t work well right now? The answers to what doesn’t work well will drive how we move forward.
Is it failed deployments? If so, then is this bad code? Let’s test more. If it’s people not always running all the scripts, or the right scripts, then we need a test environment where we actually run scripts to see if they work.
Do you lose track of what changes developers have made? Let’s make sure we have version control as a funnel to capture all changes, and from there we can look at manual or automated deployments of what’s changed.
Are you struggling to get code that meets the requirements? Again, testing can help, but maybe you need to structure better communications between developers and business customers, and make small changes that your customers can actually see and test.
If you can’t keep up with requests from developers, then maybe this isn’t about database development and this is more of a staffing issue.
The devil is in the details, and to make a successful leap to a smooth DevOps process for database development, we want to adapt to your environment. Let’s address the pain, and minimize the changes we need to make.
There isn’t a once size fits all. This is a set of tools and principles we want to use, but in a way that works best for your situation.
Pingback: T-SQL Tuesday #152 – Round up – Deb the DBA