Why is my website so popular in South Korea?

July 18, 2006 10:28 AM

Why in hell is my site getting so much traffic from South Korea?

It's not that I don't like visitors from other countries, but it seems weird as hell that half of my visitors comes from South Korea. At first, I thought it was some kind of trackback spam, but then I wondered if people were searching for information about a payment system called giro and got me instead. How can I tell which is which?

Now, this wouldn't be a problem, except I'm gearing up to podcast audio from my site, and South Korea's sucking up a lot of bandwidth (48.59 MB vs. 103.46 MB from the US in June '06). Is there an easy way to divert all of this traffic to a stripped down site? I'm pretty sure I don't have access to Apache, but I can futz with .htaccess to my heart's content.

posted by RakDaddy to computers & internet (16 comments total)

I do think the "self-link" is relevant information to answer the question though.

Don't you have webstats that could tell you what keywords are leading people to your site, and what referring sites are giving you traffic?

That's where I would start.
posted by visual mechanic at 10:39 AM on July 18, 2006

South Korea is 5th in my website stats. I believe, in my case, it's wiki spam or comment spam. Did you once have comments enabled on every blog entry? I remember when I had a blog I was getting hundreds of comment spams a day. I disabled comments and my logs showed IPs trying to access my comments.php. I deleted the comments.php and for months after, they still tried to access it.
posted by daninnj at 10:50 AM on July 18, 2006

If you start denying IPs in your .htaccess, you'll find yourself never stopping. Too many proxies, DHCP-fed dialup banks, and the like, make blacklisting IP addresses (or even small ranges of address) a thankless job.

Besides, who's to say you aren't cutting on legit visitors in the same brushstroke?
posted by thanotopsis at 10:56 AM on July 18, 2006

Do you have access to referer logs? You can see if these visitors are being referred from searches and for what terms.
posted by voidcontext at 11:01 AM on July 18, 2006

I believe (but don't have the time to back it up with a link) that South Korea has the most houses on broadband, at the highest speeds, in the world. A couple of friends lived there for a few years a few years ago and were shocked at the sad state of affairs in terms of connectivity when they moved back to a large US metropolitan area.

If you had all the bandwidth you wanted and were interested in expanding your horizons I'm sure you'd do the same. I'm happy when it takes less than a couple of minutes to download my email on my fastest available in my neighborhood 256K DSL.
posted by togdon at 11:01 AM on July 18, 2006

Sorry if that felt like a self-linky, but I figured the Korean giro link wouldn't make much sense without the domain name. Next time, name without link. Mea culpa.

daninnj, all comments are turned off by default. And I have SpamKarma 2 running, so a lot of stuff gets trapped and killed.

visual mechanic, "giro" is the top search term. And while there are some referring sites with ".kr", most of them are from feelor.com, which seems to be a Chinese site.
posted by RakDaddy at 11:06 AM on July 18, 2006

You can definitely block (deny) various IP ranges with .htaccess, but I'm not sure an IP Range redirect is possible with vanilla .htaccess unless your server has been set up with the Apache rewrite module. If it has, you could use the following solution in your .htaccess file:

RewriteEngine On
RewriteCond %{REMOTE_ADDR} ^123\.45\.6\.$
RewriteRule .* http://where_you_want_to_redirect.com [R,L]

This will redirect the block 123.15.6.* to http://where_you_want_to_redirect.com.

If they have a limited block, lets say 123.45.6.25 through 123.45.6.46

RewriteCond %{REMOTE_ADDR} ^123\.45\.6\.(2[5-9]|3[0-9]|4[0-6])$
RewriteRule .* http://where_you_want_to_redirect.com [R,L]

Google "apache rewrite module" for a ton of more resources.

(on preview)

Thanotopsis is right. It's going to be some work and you risk blocking out legit visitors in South Korea. A possible solution to that problem would be to use the index to redirect everyone to one page, Koreans to another, then put an explanation and link on the stripped down page for those poor legit souls who really want to see the real site. It'd be a bit of a juggle.

However, if bandwidth is the main concern, I can't think of a better answer to your question.
posted by empyrean at 11:07 AM on July 18, 2006

To restate my original question: is there an easy way to divert all of this traffic to a stripped down site?

Thank you, empyrean. I'll look into the rewrite module.
posted by RakDaddy at 11:11 AM on July 18, 2006

If they're coming to your site to either to comment-spam or as the result of some wayward search results, then would it be safe to assume that they won't be downloading your podcasts?
posted by diastematic at 11:30 AM on July 18, 2006

I think mod_rewrite IS the easiest way to divert this traffic. And while writing instructions for mod_rewrite is not Jedi material, I'm not sure I would consider it easy. It's an extremely powerful tool, though - it's worth learning.
posted by dammitjim at 11:46 AM on July 18, 2006

Can you see whether there is one particular IP that is sucking down most of the bandwidth?
posted by smackfu at 12:10 PM on July 18, 2006

you risk blocking out legit visitors in South Korea

Honestly, unless there is some reason you specifically want Korean visitors, I really wouldn't worry about it.

Also, 50MB/month is literally nothing; it's the background radiation of the Internet. If it's causing you problems, you should re-evaluate your web hosting.

Also, Koreans (spambots or people) will be much less likely to actually download English podcasts.
posted by trevyn at 12:28 PM on July 18, 2006

[removed link from post, put it in your profile if you want people to look at it]
posted by jessamyn at 12:40 PM on July 18, 2006

You can host creative commons licensed podcasts for free at Internet Archive.
posted by Doctor Barnett at 12:52 PM on July 18, 2006

might it just be legit mistakes by people who are looking for the banking site www.giro.or.kr?

can you tell if they stay on your site or if the traffic is just someone loading up the main page, going 'oops' and then surfing elsewhere?
posted by sergeant sandwich at 12:52 PM on July 18, 2006

Holy crap, sergeant, that must be it. Sandwiches are on me.

So, okay. I can ignore the whole thing. But now the geek in me wants to try out mod_rewrites and make a text only page with a link to giro.or.kr (and to me, if they really want to hear my melodious voice). So, it's off to tinker. Thanks for the advice and information, everyone!
posted by RakDaddy at 1:20 PM on July 18, 2006