Tell HN: took full slashdot effect to my 4$/mo server no problems

14 points by jpirkola ↗ HN
I am using cheap shared no traffic limits web server and I got a link through to slashdot, which was quite an exciting thing to happen. I got 20k pageviews in a day, and 15k unique visitors. The web server worked just fine with no evident lags - or complaints from the hosting company. So it can be cheap, and scalable. Here's the slashdot article: http://developers.slashdot.org/article.pl?sid=09/03/16/192211

25 comments

[ 2.1 ms ] story [ 67.0 ms ] thread
/. is not what it used to be. I have contributed more to HN in the last month than I have to slashdot in the last 4 years.
It's not the hip new place any more, for sure. On the other hand, they seem to have done an ok job of stemming the downward slide, something that digg and reddit are really struggling with.
Slashdot is what it always was. They have maintained their identity in spite of all the new trends, and they will continue to maintain their audience. Digg, Reddit, HN.News are different from Slashdot.
I think he meant in terms of total traffic.
The traffic they send is not.
at least according to quantcast, their traffic numbers have been steadily going down: http://www.quantcast.com/slashdot.org#traffic
That begs with my analysis. MyHosting.com hosted my blog when it got posted to reddit and /. back in Aug. 2005, and I got something like 40,000 hits the first day, with over 18+K uniques, last time I checked.

Then again, how much traffic I really received I'll never know. If you're doing more than 50 connections/sec, they shut down your account so their other clients "don't suffer".

I promptly moved. And I don't recommend them.

HN has Alexa rank of 7000, while slashdot has about 1000. When I look at HN Alexa rank, a big jump happened in the last month. Anyone knows why? I like HN myself, the philosophy suits me better...
Can you provide some more details? What kind of caching do you use? Some in the cms or on the server? Do you know if your hosting provider is using squid?
I am using stock Wordpress 2.7.1 (2.7.0 at the time of slashdot effect). I have not enabled any caching plugins at Wordpress. All the pages are generated with php, so I suspect it is impossible to effectively use squid? Here's the website itself: http://www.cybertechnews.org
Shared hosting isn't as bad as people make it out to be — very few people use many resources, so your site could have had a massive share of a 8-core box or something.

Still don't like it though, but my main requirement is root access. :)

Is your host the same one running those $4.84 ads on your site?

In any case, details on how you did it would be most interesting. What kind of changes(if any) did you make to your "hot" page?

Yes, you can actually see "powered by" link at the bottom of the page too. If you go to about page, you can find a coupon code to get even lower price (and I will get a lower affiliate slice). I didn't do any changes to the page, I just thought that it was a good story and I submitted it to slashdot. They have rejected dozens of my submissions so I didn't expect anything...
To be quite honest, 20k pageviews is no more than any server worth while should be able to handle. That's less than a pageview per second. Even peak times of 10 or 50 times that should not pose any problems...
Yes and I am sure they would come in nice a regularly like that :-)

The top story on slashdot is there for about a hour so I would expect that they would come in as big spike at the start and then drop off.

If you would have hit the frontpage regurarly then they will come regurarly...otherwise the number will fall. It's like a commercial: you need to push it more than once to have an effect.
Slashdot traffic isn't quite what it used to be. That said, it's cool that your shared hosting service was able to keep your WP powered blog up. Wordpress can be notoriously resource heavy.

Give your hosting company a heads up that you got slashdotted and they were more than capable of handling it and send them your thanks. They should appreciate it! :)

Not to rain on your parade, but developers.slashdot is not the "full" slashdot effect. The full slashdot effect (from personal experience getting http://inter-sections.net/2007/11/13/how-to-recognise-a-good... slashdotted) happens when you're on the front page of slashdot, and results in about 100-150k visitors from slashdot itself, accompanied by another 20-50k visitors from other sites that pick up the story. If you get front-page-slashdotted, you'll probably also get dugg and redditted at the same time.

For the record, danieltenner.com has had 50k visitors in a day from just reddit + HN...

The problem also usually comes up when comments start to add up (another reason why disqus rocks). http://inter-sections.net/2007/11/13/how-to-recognise-a-good... had a couple of hundred comments within a day, and since Wordpress (it was on WP at the time) didn't cache them, the page load got excruciating.

That's not a problem for http://danieltenner.com anymore, since comments are now served by disqus, and the pages themselves are static html. I suspect I could take several "full" slashdottings at the same time now and still stand. Nginx is pretty good at serving static files.

The article was in the slashdot frontpage. It got only ~80 comments so maybe it just wasn't so popular. I am still getting hits from there and all round from the sites that picked it up.
That is weird. A few months back, I got suspended from my shared hosting account after upgrading to Wordpress 2.7X and I was getting only around 600 unique visits a day... The hosts says I was already using more than 90% of their resources...
Wordpress is a known CPU hog.
Out of curiosity how much MB/GB bandwidth did this total too even though your host doesn't limit you?
The first slashdot day was 2.5 Gigabytes and the second 2Gigabytes of traffic. Then the traffic gradually slowed down, but ended up at double from the previous.
wow, It's surprising how small that actually is in reality
I got 20k unique visitors from stumbleupon in less than 24 hours on an equally cheap host (running wp, no caching, image-heavy page) with no problems.

Shame they totally screwed up a few months down the line and I was forced to move host.