Bad Advice

It’s no secret that I’ve been fairly active on a couple of SQL forums for a couple of years. In that time I’ve seen all sorts of behaviour on forums, good and bad. By this point, most of that just doesn’t bother me any longer. There is one thing however that still gets me angry when I see it. Blatently bad advice.

Now, I’m not talking about first-attempt solutions that solve half of the query problem, I’m not talking about honest mistakes and I’m not talking about attempted answers to questions so vague they’re near-impossible to understand. I’m talking about advice that’s so bad it’s dangerous. I’m talking about things like this:

Q: My transaction log’s very large. What can I do to fix this?
A: Stop SQL, find the ldf file, delete it and then start SQL

Q: Dropping a clustered index on a large table takes a long time using drop index. Is there a faster way?
A: Run sp_configure “allow updates, 1 and then run delete from sysindexes where id = OBJECT_ID(‘MyTableName’) AND indid = 1

Ouch!

There’s two main problems with bad advice.

Firstly, the person asking possibly doesn’t understand enought to realise the advice is risky, and if they follow it without testing they could end up in a much worse situation than they were. Dependiong on the circumstances they may end up in trouble with their boss, they may even end up getting fired. Whether they realise the advice is bad before or after trying it out, it’ll erode their faith both in the forum and in the person who provided the information. That leads to the the second problem.

The second problem is damaged credibility and reputation. A good reputation is so hard to get in this industry and so easy to lose. Bad advice damages the forum’s reputation and the reputation of the person posting the advice. Brent Ozar puts it very well: “Being an expert means having credibility. It doesn’t matter how much you know if people don’t trust your answers.”

Then, of course, some other person on the forum has to come along and correct the bad advice and explain why it’s bad so that people who find the thread via google won’t think it’s useful

Bottom line, bad advice helps no one and hurts a number of people.

</soapbox>

10 Comments

  1. Mike Walsh

    Here Here! This is the week of posting about doing silly things without testing first apparently.

    Great post with excellent advice. Always strive to give good advice, stay silent if you don’t know. On the flip side, don’t take everything on the internets as gospel. Test it, try it and see what happens. If you don’t understand it, learn it before doing it.

    I have seen plenty of examples about the first example but not the second. Scary 🙂

    Reply
  2. Gail

    It’s funny how several people can post about the same thing around the same time without any discussion first.

    As for that second piece of ‘advice’ by the time I saw it, the OP had already replied with ‘Thanks, I’ll try that’. I do so hope they had a backup.

    Reply
  3. Brent Ozar

    HAHAHA, wow, I love the delete-LDF answer. I know the job market’s tough in some cities, and maybe they’re trying to sabotage somebody else in order to take their job! Heh.

    Reply
  4. Jack D Corbett

    I just read one where I person asked for a time bound trigger and someone suggested that they use WAITFOR in the trigger. AHHHHHHHH!

    Reply
  5. Pingback: The Rambling DBA: Jonathan Kehayias : Online Credibility, How to build it and how to lose it in an instant

  6. Travis Alltop

    I couldn’t agree more Kimberly! Giving bad advice to someone, like your 2 exanples atate, is kind of like giving a blunt pocket knife to someone and telling them to go cut down the tree in their own front yard. Not only are they approaching this the wrong way with little fore-thought, but they could be causing all kinds of unseen trouble for the person that might axtually listen to bad advice like this, particularly if they had never done this before and just assumed the person that told them this actually knew what they were doing…. Things like this can ruin your rep as a DBA real fast as you say, and can actually get someone else fired over it toom which I think is even worse. But then again, after being in this business for almost twenty five years myself and knowing just how doggy-dog things can get in most IT shops, you would be surprised at how many times I have seen people purposely send people out on wild goose chases looking for left-handed monkey wrenches just to make them look like fools in front of their peers or superiors simply because they either didn’t personally like them or because they were really after their job. It’s rough out there people in a lot of IT shops and you have to take some responsibility upon yourself to ensure that you just don’t blinfly follow every piece of SQL advice you get, regardless of the “guru” it came from. A little common sense and research before acting upon “any” advice from a so called “expert” is the rule of the day in my opinion…..

    Reply
  7. Gail

    Kimberly?

    Reply
  8. Travis Alltop

    Sorry Gail, I was reading one of Kimberly Tripp’s blogs yesterday and had her name on the brain. 🙂 My point is here is that the focus should not be on how a DBA should seek outside advice to do something, which leads them into getting into this type of bad situation. Rather, the focus should be on whether they should be doing it in the first place… Always think before seeking any advice that could risk your db integrity. “Should I even be doing this right now????”

    Reply
  9. X002548

    ok then

    Glad I didn’t get mentioned for bad advice

    Great Stuff “Kim”

    Reply
  10. Pingback: Shrinking Is A Popular Topic | SQL Server Blog - StraightPath Solutions

Leave a Comment

Your email address will not be published. Required fields are marked *