A perfect storm of traffic
I’ve been busy! Project B-2 Bomber went through a “perfect storm” of traffic. I had 4 all-time record breaking days in a row – meaning, every day broke a whole new record for single day unique visitors.
- Day 1: one post gets picked up by a blogger whose link sends hundreds.
- Day 2: same post gets picked up by a number of smaller bloggers, who collectively send hundreds more.
- Day 3: same post, plus several others, including several very old ones, get picked up by Stumble Upon users, who send thousands
- Day 4: an industry news source picks up yet another post and sends a few thousand – while the Stumble Upon/blogger thing is still going
- Day 5: traffic slows down somewhat, but I gain some really valuable inbounds
My feed subscriber record got broken, too. This all resulted in… not a Digg effect, but a nice steady wave of traffic, complete with some new commenters – most of whom had something useful to say and will hopefully stick around as regular readers (as opposed to the time a snarky forum sent us 25,000 people in a single day, most of them leaving nonsense comments just to see if they could overrun the servers – which they did not! Let’s hear it for DownTownHost!).
Project Mai Tai is also breaking some records in the feed sub department. Traffic is increasing a little bit, and I’m gaining inbounds that’ll serve me well down the line. By the way, am I the only one who thinks FeedBurner is way low on some of its numbers? I’m the first to say when my sites aren’t doing well, or when I’m surprised by how well they’re doing, but this is the first time both sites made only a modest gain out of the sort of traffic that has, in the past, given me much better feed numbers? Looking at Feedburner’s specifics, I see they don’t count the same services every day, which might mean they rely on services to report accurately, and therefore can only be as accurate as the reporting they get. Ah, well – it still seems to work as far as comparing two sites goes.
My next quest: figure out how to Digg-proof Wordpress without using caches. I’m sick to death of caching software. Yes, it’s awesome – the server doesn’t even burp – but it also caches ads, which is just not good. It also causes post edits and style changes not to show up quickly unless you manually delete the cache – which is not something my contributors on B-2 can do. And finally, it can cause people to miss each other’s comments and interrupt the flow of good discussions. I found this article on speeding up Wordpress without a cache, but the only way I can know if it works is to disable WP-Super-Cache, implement that solution, then find out the hard way if it doesn’t work the next time I get tons of traffic. Anybody know anything about this? Will that guy’s solution Digg-proof a WP install?
Related posts:
Categories:
Tags: |
July 23rd, 2008 at 9:42 am
Hi Sapphire,
I got your ping.
WP Super Cache may or may not Digg-proof a WP install because it all depends on the number of HTTP connections. The cache may speed up the loading, but will it speed up the disconnects?
Caching is a good idea, but I don’t like the way WP Super Cache does it. I use Hyper Cache, modified with some lines to make 2 cache files; one for search visitors and one for non-searchers, just like my blog without the cache on. Is it Digg-proof? Again, the same question applies.
It depends on how the host has it set up. Most shared servers choke at 500 HTTP connections. Every image, Javascript, and CSS file loaded with the blog counts as a connection. It doesn’t take much to reach 500. I once hit the home page and had 500 connections in the first minute.
The only way to prevent ads from being cached is to make sure they’re served with JavaScript. AdSense, for example, is not cached.
July 23rd, 2008 at 12:45 pm
Dang, you’re quick!
I started responding here, then decided I might as well try Hypercache. I deleted SuperCache (I think…. it stashes folders and files like squirrels stash nuts, but I think I got it all). So far, all I can judge about hypercache is how quickly pages load, and I don’t have a tech savvy way of testing that. It feels quicker to me.
I reduced server calls greatly by changing a lot of PHP calls in my themes to HTML links and moving some theme images to Flickr. I agree that it’s essential to reduce the calls, no matter how you cache ‘em.
Thanks for the help! Hopefully this will work better for me than SuperCache did.
July 23rd, 2008 at 3:38 pm
I’m glad you like it. Feel free to stop by and ask for theme/hyper cache modifications if you need something that doesn’t exist.
August 10th, 2008 at 7:30 pm
Why are cached ads a bad thing? Can you lose ad revenue from that?
August 10th, 2008 at 9:05 pm
Yeah, Mike – if the ads are paid per impression. Caching would cause it to register fewer impressions.