Less than 24 hours ago I blogged about a veritable nightmare I had with CHECKDB and the resulting findings that it wasn’t actually a corruption problem but instead a new issue in a third party tool that we use on our production servers. This little problem actually answered a long standing question that was posed to me by Andy Warren(blog/twitter) when I presented a session for the OPASS user group earlier this year. The question was basically “For the size of your network, what has it actually done for you?”
Since Andy posed that question, I actually have made a lot of improvements in my networking. I use a number a tools for networking, various Forums and News Groups, Twitter, Linked In, Facebook, and not the least of which happens to be my blog on SQLBlog. Last night my social network saved me from continuing to troubleshoot what turns out to be a non-issue. Less than 12 hours after making my post, A tale of CHECKDB failures cause by 3rd party file-system drivers, my SQLBlog network and exposure worked to my benefit.
One of the great features of the Web 2.0 platform is that you can literally receive real time updates of information about any topic that you want to know about on the web. One of my personal favorite things out right now is Google Alerts, which give you near real time notifications of new web content containing key words that you specify. A lot of businesses have alerts setup for their business name, product scope, or area of expertise, so that they can track what information is currently available online regarding their area of focus. This makes a lot of sense to me. I personally use Google Alerts extensively, and I have multiple alerts defined for things that interest me like SQLCLR, SQL CLR, Extended Events, XEvents, Service Broker, and last but definitely not least, my name, Jonathan Kehayias.
Why do I do this? Well I like to know when new information is available about the technologies that I am passionate about, but more importantly, I like to know when/if my name is being used publicly. I’ll admit it generates some additional emails at times, for example, this and every blog post I write generates an email to me, but I have also been able to see where things I said are being questioned, and where I am being quoted/misquoted, or referenced. A lot of Web 2.0 savvy businesses are also keeping tabs on their brand and market area’s to see where things are being said. For example, when I woke up this morning, a competing company to Diskeeper, PerfectDisk, had already tweeted about my issues with Diskeeper 2010 causing CHECKDB issues. This information was less than 6 hours old at the point that they had picked up on it.
What really amazed me though was that less than 10 hours after posting my blog post about my experience last night, I had a comment from Michael Materie, (if he reads this I’d be interested in an email through my blog), at Diskeeper not only confirming what I had found but also stating that the problem had been identified, fixed and was going to field testing immediately. We were in the process of starting a support case with Diskeeper when the comment came through,so we hadn’t even made direct contact with the company yet.
What scares me the most about this entire scenario is that the new version of Diskeeper had been in public beta, and released and none of their customers caught this bug, which wasn’t a problem, but definitely looked like one and cost the administrators at my company and me a few hours of stress and lost sleep. Let me be clear about one thing, the bug in the product is a problem, but the bigger problem here is that no one testing this had run CHECKDB in SQL over a test period of more than a couple of months. As well documented as the importance of CHECKDB is, people are still not running regular consistency checks of their SQL Server databases. This is so disheartening to think about especially given the vast range of SQL Server implementation that we are seeing today. If you aren’t running CHECKDB weekly, or at the very least bi-weekly, you are leaving yourself open to problems that are very preventable. The newest version of Diskeeper was in my environment less than a week before this problem popped up.
One thing I’d like to address is why I posted the blog I posted. I didn’t do it to bad mouth the Diskeeper product, I did it to immediately document the problem I encountered, with error messages, and resolution, so that someone else in the same scenario but without the network of people that I have available to me wouldn’t waste time and possibly money on support calls to Microsoft, VMware or hardware providers, chasing something that is the last thing in the world that you would ever think about causing a problem. One of the side effects of this was a little egg on the face of Diskeeper, but I think that their immediate response to the problem, and their professionalism has been outstanding, and I will continue to recommend the product for online defragmentation of disk arrays that contain SQL Server Database files when appropriate.
I know that this post has strayed from its original intent, but one of the things that stood out the most to me in this event is the power of a good network, as well as the power of a good online presence. One of the things that I like most about Andy Warren is that he constantly challenges my normal way of thinking. He really likes to ask the tough questions. I’ve really paid a lot of attention to Networking based on what Andy writes about and talks about. One thing I can say about Andy is he talks the talk and walks the walk. Good Networking is critical to Andy. When Andy asked me what my network had done to help me, I didn’t have a good answer. Less than a year later I can definitely say my network has done a lot for me.