I don't understand how this got to the front page. Would anybody upvote an article about how adding indexes to a table in MySql improved performance?

I totally agree with you. What is the lesson here ? You need indexes to speed up queries in any structured data storage. Uh, great but I am worried about the glipho backend then. All this to get a bit of advertising.

I partially agree,

I don't see how one would not find this obvious after reading the mongo docs which explicitly say

> Indexes enhance query performance, often dramatically ...

I think the point here is that many people don't realise indexes matter so much in MongoDB. Their documentation doesn't go out the way to push indexes as a must. That coupled with the ease at which you can just start using it compared to say, MySQL, means that people find it interesting.

I will point out that while I wrote this, it was more to try and ensure other people don't make the same mistakes.

Exactly,the article doesn't really have anything much to say. Overhyped,headlines all over again.

I might be nitpicking, but your blog's permanent and space taking headers make the user experience very unwelcoming, and make the whole site feel unwieldy -- slow scrolling and a good amount of space being taken over for no reason.

Hi, thanks for the feedback! It's not my blog actually, we are a startup (hence the top bar for users), we will take the feedback on board, maybe we got a little excited with the size of the topbar :)

Worse than that: from my end on Chrome (beta) running in Ubuntu, the bars appear to stay in their original location when scrolling which makes them cover content and generally bug out. I could not read the article at all!

I think it's the combination of the top and bottom bar that make the site feel a little weird for me.

Hi there!

I've been a Glipho user since its recent launch... and I'd definitely say that you are being nitpicking :P

This is a newborn platform. Even though, it has a lot of things to do, it is a nice approach. You can't expect it to be a perfect fit in every way yet, can you?

Nevertheless, this post was about MongoDB, why focusing on usability?

Article could be made useful by showing a benchmark of indexes vs no-indexes, and showing the readers how the indexes were created. At least then we can say it's educational and has substance. Otherwise I agree this is substance-less. It's common sense that any database is slow if used incorrectly.

Point taken! We'll make sure next time we bring something more quantifiable and more stats. thank you!

I am convinced that in any production database cluster If you are given the option you should disable table scans. MongoDB gives you that option, you should use it.