I’m hosting a new webinar on the DLM tools from Redgate Software and database DevOps. Tomorrow, Tues, Jan 17, at noon EST, Arneh Eskandari, a great sales engineer, will be demoing the tools and techniques to deploy changes.
You can register here and I’ll talk to you on tomorrow.
Redgate produces some great software, and I’m glad that I get to use all of them as a developer and DBA. I use many of the tools each week, some daily, but I constantly find that there are plenty of features, tips, and tricks that I don’t know about.
At SQL in the City last year, I watched Tom Austin present a session where I learned a few things that will make me more productive.
The one that I need to start incorporating into my work is looking for all the references to an object with SQL Search. A quick check means that I won’t commit changes without considering all of the other objects impacting by development,
Watch all the tips below.
There’s a saying in the data world: garbage in, garbage out. We use that when we can’t get good information from our database because the data we’ve stored isn’t as useful as we would like. That’s a problem, and it’s one reason why data professionals want to spend time thinking about the data we need to collect and how to store it. We want to be sure that we’ve at least made an effort to collect useful data that someone will use.
We sometimes have the data we need, but still struggle to use it effectively. I think this is an area where machine learning and similar technologies may help in the future, but there is a lot of work to be done to allow most of us to take advantage of those tools. In the meantime, many of us make do with basic T-SQL to perform data analysis, generate reports, and provide the answers to questions. When we do so, it’s important that our queries actually work correctly to answer the questions we need.
I don’t want this to be a political discussion, and I would appreciate that any comments be limited to the technical subject. I ran across a piece about a failure of the US government in determining the status of people being checked for immigration status. The interesting quote in this article was “… officials blamed computer code for the problem.” Leaving aside the implications in this case, the idea that computer code, likely some sort of query code, is not working as expected, querying the correct data, or isn’t being used properly is disturbing.
I’ve run across quite a few stories like this from various consultants that were called in to help organizations, only to find out the queries that had been used for long periods of time were incorrect. They didn’t filter appropriately, didn’t convert or aggregate data as intended, or didn’t even query the correct data.
We use databases and queries extensively in today’s world, and the growth is only going to increase. As much as I like the idea of DevOps and more frequent deployments, I also want higher quality for our software. This means that we need to ensure that our queries actually work as intended against databases. Code reviews, independent checks, using known data sets that evolve and include edge cases of data are all ways we can work to ensure we are actually writing the correct queries for our data. Above all, we need to be sure we are using test of some sort, preferably unit tests, that ensure the queries are actually the ones we want.
Steve Jones
Listen to the MP3 Audio ( 3.8MB) podcast or subscribe to the feed at iTunes and Libsyn.
When I want to upgrade hardware, I usually turn to Glenn Berry for advice. He’s the person I know that really watches, examines, and evaluates computer hardware. He gives good advice and upgrades often, so he actually uses the latest hardware.
One of the upsides of his hobby is that he is regularly replacing hardware. That means he regularly has older hardware. A few times I’ve asked for advice, I’ve actually ended up buying the previous generation from Glenn because he’s already upgraded. That works out well for me, since I don’t look to be on the latest and greatest.
Recently we had lunch and Glenn offered me his old Radeon R9 video card. It was ranked quite a bit higher than the GeForce GT 740 I bought last year. I didn’t need a replacement, but the price was right, and maybe I’d impress my kid a bit with the upgrade.
I went to lunch, and got a nice, geeky gift bag.
Inside was not one, but two video cards. Apparently Glenn’s MicroCenter Outlet was having a special.
The cards are on my desk as I need to some time to open up my case, replace cards, and then ensure everything comes back up correctly. I’ll need drivers, and no pressure to complete things quickly.
Hopefully I won’t need a new power supply, though I am tempted to get another monitor with all this horsepower and give Twitter and Spotify their own displays. 
