Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
WordPress.com Goes Down, Takes Millions of Blogs With It (mashable.com)
44 points by dwynings on June 11, 2010 | hide | past | favorite | 29 comments


We think central hubs are great -- till they fail. We are always shocked when they fail, but shouldn't be. We expect perfection from a single entity. We can't.

Maybe it's time to change our focus.


The bonus of a single entity is that they can afford to hire experts to solve various problems that it wouldn't be feasible for smaller groups to hire/solve. (e.g. Lots of people prefer to have their email done by a provider rather than spend untold amounts of time configuring exim/postfix/qmail/sendmail/etc, in addition to setting up spam filtering, managing blacklists/whitelists/greylists, solving issues when your domain and/or IP ends up on one or more blacklists, etc)

Not that I'm necessarily a fan of significantly centralizing things, but for some things I can see the appeal of off-loading the complexity onto someone else.


There's an upside of central hubs too:

http://www.expatsoftware.com/articles/2008/02/amazon-goes-do...

When they fail, it's not your problem. Better still, it fixes itself quickly, without you having to get out of bed at 3am.


It's an interesting risk trade off. The central hub is complex but well resourced so able to solve problems quicky (hopefully), self hosting is perhaps a simpler overall solution as it is smaller scale (so less chance of a failure?) but you probably don't have the same man power to deal with a crisis.

Which is more likely to suffer an outage? Which is more likely to suffer a long outage? Which is more likely to suffer a fatal outage?


It's an interesting problem. Here's my (current) example of a problem with central hubs.

I had shared hosting with a provider; mistakes and bad management lead them to, basically, kill my account for most of a day - including all my sites.

Which sucked.

Since then I've been moving to my own VPS's - currently using two shares from Gandi to host sites, provide my own DNS and so on. I actually tried this a while ago and it got bogged down in complexity; now powerdns, nginx et al are much much simpler and it's only taken me a couple of days to roll out.

To be honest; I think with cloud services becoming so much cheaper we will see more and more of this - with us relying on "hubs" for our infrastructure rather than our front facing sites.


Totally agree. It's incredibly easy to criticize a central hub for any downtime whatsoever because the impact is so large and visible.

But your criticism is completely baseless if the average downtime on a central hub is less than the average of a self-hosted solution.


When your self-hosted blog goes down, it's usually your own fault. I guess that's probably more acceptable to hackers than when it's somebody else's fault, and all you can do is sit and stare at generically worded status updates on Twitter.


True, but to your users, it doesn't matter whose fault it is. What matters is how quickly it gets solved. And I'm only guessing, but these things probably get solved quicker with central hubs.

Frustrating, but that's the way it is.


Change our focus away from central hubs, or change our focus away from the expectation of perfection?


Wordpress being down is an inconvenience, gmail being down is a pain in the ass. I expect better up time out of my email service than out of a blogging service.


I wonder if the continuous development/deployment model that they use is somehow responsible.


"The cause of the outage was a very unfortunate code change that overwrote some key options in the options table for a number of blogs." TC

http://techcrunch.com/2010/06/10/wordpress-gives-us-the-vip-...


Wordpress is one of the largest sites on the web now - I can't believe they have a single point of failure


Could be an unpredicted cascading failure. 24x7 is very hard.

I am glad air-traffic control does not rely on Wordpress ;-)


If it was server storage related, then we don't mind helping them out =)


I have been in this situation with high profile sites before. When there is downtime that is very public like this incident, you get a flood of emails and calls from service providers offering to help. This can actually often work, a relationship I established with a particular hosting provider (who I introduced to numerous other people as well) started out during a downtime incident when an executive of the company sent an email. If you are a startup that offers services, I would recommend taking advantage of these opportunities to get new business (put on a good pitch at a time when the potential client is vulnerable).


Ironically, I am trying to go to your site to see what you do and am getting the following error:

  Error

  Request-URI Too Large

  The requested URL /... is too large to process.


And when I try to view http://moolave.tk/ in Firefox I just get a message saying:

This content cannot be displayed in a frame. To protect your security, the publisher of this content does not allow it to be displayed in a frame.


It works for me but seems to be a weird structure of a regular site hooking in a google sites frame which bootstraps a wix flash page which frames the actual site content.


Yes, we designed the website from the Wix Flash platform and published it on Google Sites. We still have to make some changes as the right-hand side is taking some space on high-resolution screens.


I also got NoScript's cross-site scripting filter kicking in which is never a good thing...


Thanks for the input. For some reason it works on Chrome and Firefox on my side. I will check and see what I find.


Thanks Jeff. It might have something to do with Google running slow (with Gmail and other applications having the same challenge). I will check and see what's causing it.


It could be one of many problems, but I checked a similar thread reporting the same issue; only this time the issue was addressed under Gmail (still Google)

http://www.google.com/support/forum/p/gmail/thread?tid=35602...


They use Amazon S3 for storage in the past, not sure if it is still valid for now.


Then I hope we can help them if given the opportunity.

The site is working so far. Thanks for the feedback, folks.

http://moolave.tk/


Yet another great reason to self host


...if you have better uptime than wordpress.com.


So kind of a long-tail problem, then.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: