1. When evaluating performance of a query, logical reads matter more than duration.
2. If you think that Reports in Reporting Services are boring, you need to learn to use F4.
3. SQL Server has its own, trimmed down version of the CLR.
4. Don’t be afraid to be uncomfortable.
5. Document before, during, and after a process.
6. Before investigating another 3rd party reporting tool, download Report Builder and give it a whirl.

Unfortunately, these points don’t even begin to cover everything there was to learn today. I want to thank Randy Knight (blog|@randy_knight), Doug Lane (blog|@douglane4), Mitch Bottel (blog|@SacSQLDude), and TJay Belt (blog|@tjaybelt) for their excellent presentations. Finally, thanks to all the sponsors for making these events possible. For those of you that have a SQLSaturday coming to your area, I strongly recommend you attend.

Related posts:

1. Utilize the Languages You Have Learned
2. Valuable Lessons Learned
3. Hey Quest! Thanks for the Swag!

The other day, I went to the local Ikea for the first time. The girl at the door gave me a pamphlet with a map on it that showed the route one would take through the showroom. However, on the map there were handy “shortcuts” marked for people like me, who already know what they want. I wonder if I spent more time trying to find the shortcuts and navigate through all the displays than if I had just followed the route indicated on the map.

Choosing to ignore process is choosing to waste time

In software development, there is always the temptation to take “shortcuts.” Pressing timelines force developers and managers to ask the questions like “what if we just chose to not follow the process this one time?” It’s times like that when I can’t help but wonder if we spend more time trying to shortcut the process than we originally would have just doing things using the defined process.

One shortcut leads to another

This week, I’ve felt the pain. In trying to get our product out, with a shortened deployment schedule, we’ve chosen to take shortcuts in the process. At each decision point, I’ve tried to remind the decision makers that we’ll likely end up redoing things because we’re choosing to abandon process. So far we’ve had that discussion 3 or 4 times over the last few days. Each time is the direct result of choosing the shortcut in an earlier decision.

Shortcut Warning: Here Be Dragons

My advice to any who are considering shortcuts, remember that shortcuts usually lead to wasted time and poor quality. In map makers terms, shortcuts should be labeled “hic sunt dracones.”

Related posts:

1. Unit Testing Will Save Your Bacon
2. Database Unit Test Structure
3. Understand Your Customer’s Viewpoint

1. GUIDs as primary keys
2. Redundant data tables
3. Triggers, triggers, everywhere
4. Unused tables
5. “One procedure to rule them all” mentality
6. Objects that follow no known naming convention
7. Business logic hidden in functions five levels deep in nested procedures
8. The quote, “Just get it to work, we’ll worry about performance later”
9. CURSORS!

Related posts:

1. Know Thy Data – The First Rule of Data Audits
2. Fillfactor and Page Splits
3. Database Unit Test Structure

Last week, I attended the local SQL Server user group meeting and heard a good presentation on index maintenance given by Martin Miller (blog|@mmillerDBA). During the presentation, I had some questions about how the fillfactor property on an index affects when a page split occurs. I came away concerned about the answers I was given and soon realized that I had fallen victim to one of the classic blunders.

No, I didn’t start a land war in Asia. Nor did I feel that the presenter hadn’t done his best to provide me with an answer. I was trapped in a slightly less well known blunder of relying on my own false assumptions. It took a little bit of study before I realized that I had a fundamental misunderstanding of the fillfactor property.

I learned something from a false assumption…

I’m trying these days to adhere to the Igor Stravinsky model of progression by learning “through my mistakes and pursuit of false assumptions.” I came away from the meeting questioning whether I really understood fillfactor and how it affects page splits. I just couldn’t understand how maintaining an 80% fill factor would keep a page from splitting any more that maintaining a 100% fillfactor. So, when I got home that evening, I cracked open Books Online. It was then I realized my mistake and I learned something from that false assumption.

I had an incorrect understanding…

For those of you that have never heard of a fillfactor on an index, it deals with the underlying structure of an index on the disk. In the same way that data in SQL Server is stored, indexes use pages to store their information. Normally, those pages would be filled to capacity, at which point any incoming index records would cause the page to split in two half-filled pages allowing for the additional records to be inserted into one or the other of the pages. This is known as a page split. The fillfactor property allows you to indicate how full those pages should be. Essentially, you are setting the capacity of the pages in an index to include unused space for later additions to the index. Unfortunately, I had an incorrect understanding of how fillfactor is maintained on an index.

In the meeting, I started to become a little confused about when a page split occurs on an index with a fillfactor specified. I even asked the question as to if there is some algorithm that maintains the fillfactor and when the page hits some amount over the fillfactor it splits the page. I couldn’t understand how that would give you a performance gain. First, there is the algorithm itself that would add overhead to every insert, update, or delete. Second, there is the fact that the page is still going to split regardless of the fillfactor. In fact, the lower the fillfactor percentage, the more often the page will split. I couldn’t understand how maintaining free space in the page is going to keep that from happening and how that is a performance benefit.

I found the root of my misconception…

As I drove home from the meeting, I realized that I was probably wrong and had a flawed understanding of the I decided to hit the books and see what I could find out. Thank you Microsoft for Books On Line. One read through the entry on fillfactor and I found the root of my misconception concerning fillfactor and how it is maintained.

It turns out that fillfactor isn’t maintained in the way I thought. The following sentence in particular helped to clear up my misunderstanding.

“The fill-factor setting applies only when the index is created, or rebuilt.”

In other words, the fillfactor is used to set the initial fill amount of the pages in an index and then is not used again until the index is rebuilt and the data is redistributed within the pages. There is no magic algorithm that maintains the fillfactor as the index is used. Page splits are going to happen when the page hits capacity. The performance benefit comes from leaving space in a page for when inserts happen. This means you are more likely to have room for an insert rather than having to make room by splitting the page.

That being said, there are trade offs to specifying a fillfactor on an index. Given that fillfactor leaves empty space in each page for future inserts, it’s important to remember it also requires more pages to store the same amount of data. More pages means when you want to find something in the index, you have to look at more pages, which increases scan time.

If you find yourself confused about indexing, fragmentation and fillfactor, there are several good posts out there from the devsaints and there is always Books On Line.

Related posts:

1. Server Aliases, A Less than Optimal Solution
2. Understand Your Customer’s Viewpoint
3. Database Unit Test Structure

Now, I know that in the SQL community on twitter and through PASS, there are many people that reside outside of the states. However, I wonder how many of us that participate do so in a language other than English. I myself learned to speak Mandarin Chinese some 15 years ago, and used that knowledge to get a degree. I value that part of my education, yet 8+ years later and I rarely if ever get to use that language. This has to change.

I decided some time ago that I would like to make use of my language skills more than I have been. I looked into several options to increase my opportunities. Unfortunately, the majority of those options involved moving residence, changing careers, or accepting pay cuts. In many cases, it involved all three.

Last week, I again began to wonder how I could utilize these Chinese skills. Because I like what I currently do for a living, I started looking for smaller ways to use these skills without changing job or career. I found quite a few resources that involved working in a different industry. These I discounted out of hand. From the remainder, I was able to cull a few ideas listed below that you may well try yourself.

• Become a Teacher – Granted this may seem like an outright career change for some of us. However, that need not be the case. There are many non-profit organizations that need people to teach English as a Second Language in the evenings. In addition, some of those organizations are looking for professionals with language skills to teach “life skills” to immigrants who are trying to improve their employment but are better served in their native language. I remember doing this for a bit in college for a Physical Geography class. I tutored a couple of Taiwanese students, helping them learn the principles taught in the class, but using Chinese to do so. Not only was I able to help them, but I found my understanding of both Chinese and Physical Geography increased immensely. There is no better way to learn a principle than to teach it to someone else, especially if it’s done in a foreign language.
• Become a Translator – We’ve all had the experience of trying to assemble some piece of furniture and finding the instructions to be almost incomprehensible because of the poor translation. Unfortunately, this isn’t isolated to instructions for furniture assembly. Many foreign web sites, especially those for smaller companies, tend to be translated poorly. If you have skills in a certain language, perhaps you could promote yourself to some of those companies and offer your translation services. I imagine that, as more and more of us join the global marketplace, there will be an ever greater need for such localization services.
• Become a Multilingual Blogger – If you’re really convinced that a change of career or move to another country is not an option, there is always the option for using your language. Many of us are already working to reach out to our respective communities by blogging about those things we are passionate about. Whatever it is you’re passionate about, it’s a good bet that there are others who’s native tongue is not English that are just as passionate about the subject. You could consider reaching out to both communities by running a blog in both languages. Recently, I was able to install and test out a plug-in for WordPress that will allow me to run this blog in both English and Chinese. If you’re interested, check out the qTranslate plug-in.

Of course, this is not an exhaustive list of things one can do to use their non-technical skills in the workplace. I have a sense that there are a whole host of things that one can do to keep current on their foreign language skills without changing jobs or careers. If you’re interested in brushing up on your language skills, try one of the suggestions, or try to figure out your own novel approach to using your language. Regardless, while we may not see the benefits of getting a new job because of our language skills, the intellectual rigors of using another language can only improve our reasoning skills.

As always, I appreciate your comments.

Related posts:

1. Happy Thanksgiving – First!!!!
2. Valuable Lessons Learned
3. The End of the Sabbatical

In our world, viewpoints have been called many things, viewpoints, paradigms, attitudes, perceptions, etc. Whatever the name, they have a powerful influence on us as individuals and on our clients. If you doubt this, ask yourself if you’ve ever had one of those moments when you’re viewpoint on an issue was changed and you suddenly gained a greater understanding of what you were dealing with. It can literally be a life altering event given the situation. Now imagine bringing that power into your development.

The way to do this, and increase your success as a developer is to force yourself to understand the viewpoint of the user. I know that you’ve heard this before. It’s really no big secret. Yet, it is one of the things that I have seen lacking in myself and almost every programmer that I know. Nor is it an easy thing to accomplish. But the consequences of not striving for this can be dire.

Several years ago, I was having an argument with one of the smartest guys that I know about some software that we were working on. He largely worked with our government clients and I was solidly on the enterprise side of the house. The contention revolved around whether to code features for the PhDs from the government clients or for the MBAs from the corporate world. In thinking back on the argument, neither of us could understand the other’s viewpoint, and more particularly the viewpoint of the other’s clients. The end result was to keep the product moving in a “one size fits all” direction, which has led to a lot of frustration. In the end, the best we could hope for with the product was a mediocre response from the customers. Because we took that approach, the product didn’t meet the real needs of either user base and while warmly received was never really used.

Imagine the difference it would have made if we could have understood each other’s customers and split the product into two, one for each user base. Chalk that up to lessons learned.

Related posts:

1. Unit Testing Will Save Your Bacon
2. Abstraction And Encapsulation Are Our Friends, Even In The Database
3. The Form/Function Paradox

Related posts:

1. Know Thy Data – The First Rule of Data Audits
2. It’s All About The Data
3. Server Aliases, A Less than Optimal Solution

Related posts:

1. Hey Quest! Thanks for the Swag!
2. Happy Thanksgiving – First!!!!
3. The Form/Function Paradox

I’m working with a database project that has several iterations of development on it, but no unit tests. It’s a clean slate. The question I’ve started asking myself, however, is what is the best way to structure unit tests for a database project? The company I’m at tends to go off the mantra “a test for every procedure, and a procedure for every test.”

I look at this more from the standpoint that a unit test should be centered on a particular process. Usually this isn’t in conflict with the company’s maxim. Unfortunately, sometimes a process covers multiple procedures. So, what then is the best way to structure these unit tests?

For now, I’m taking the first test case and putting it in using both the company’s method and the more process-centric method. I’ll have to see which is easier to maintain and requires me to do less coding. How would you go about structuring these tests?

Related posts:

1. Unit Testing Will Save Your Bacon
2. Abstraction And Encapsulation Are Our Friends, Even In The Database
3. Server Aliases, A Less than Optimal Solution

One of the things I’ve seen mentioned to gain and retain customers, followers, people that like you, etc. is to occasionally give away goodies. Brent Ozar is doing this as we speak, offering up a couple of 120 GB hard drives that he had hanging around in his office. I’ve seen some other bloggers do this as well.

This technique isn’t the sole domain of bloggers. Companies can do this to great effect as well. Today when I got to work, I found a package on my desk from Quest software. Now, I’ve never really taken the time to use their tools, and I really hadn’t planned on doing so any time soon. Needless to say the package surprised me.

When I cracked it open, I found inside a couple of posters for SQL Server’s DMVs and a quick reference guide for 2008. I had thought about requesting them awhile back, but I had never done so. It turns out though, that I did register for and watch a presentation given for their “Pain of the Week” broadcasts.

Quest sent the posters to me just because I had registered to watch an hour of free training. In my book, that’s awesome. I recognize that it’s a little thing, but in my world, the little things matter. So, I just want to thank Quest, and they guys that I have interacted in the community that represent Quest, Brent Ozar and Kevin Kline. Keep up the good work guys, and thanks for the swag.

– Dandy

Related posts:

1. Server Aliases, A Less than Optimal Solution
2. Happy Thanksgiving – First!!!!
3. The End of the Sabbatical