Bad Web Hosting Services And The Web Infrastructure Problem

If you tried to visit Publishing 2.0 yesterday afternoon, you may have noticed it was down. And still down. And STILL down. For an hour and a half. If you tried to visit Publishing today, you may have noticed it was down AGAIN — for at least an hour.

This wasn’t the first outage with my current hosting service, An Hosting (which is owned by Midphase), but it was definitely the last straw.

So I’ve move on to yet another hosting service, one recommended by someone who runs high traffic websites, whose opinion I trust. It’s more expensive, has fewer bells and whistles, and a nearly sterling reputation, which is almost impossible to find in the cut rate commodity web hosting business. (I can’t recommend it yet until I see how it goes.)

This got me pondering the age old online question — why do all web hosting services suck? And, more interestingly, what does this imply about the state of our beloved Web’s infrastructure?

Here is a sampling of the littany of downtime woe from An Hosting:

MJ01.midphase.com

By Chris Meisinger | June 7, 2007 @ 4:59 pm

The Filesystem on this server was irreversibly corrupted. We have brought up one of our hot-spare servers and have started the restoration process. We will update this post as we have more information about the status.

Temporary Outage

By Andy Pace | June 7, 2007 @ 12:39 pm

We are currently experiencing a power problem in a remote facility. Some customers will have a loss in service during this time.

Our technicians have remedied the problem and things are beginning to come back online. I will update this post when more information is available.

Power Outage

By Andy Pace | June 6, 2007 @ 2:43 pm

We are currently experiencing a power problem in a remote facility. Some customers will have a loss in service during this time.

Our technicians have remedied the problem and things are beginning to come back online. I will update this post when more information is available.

Power problems in a “remote facility”? Just how remote is this facility?

Of course, I’m angry at An Hosting for all of the downtime, lost traffic, and missed opportunities. (The An Hosting 99.9% uptime guarantee is a lie.)

But the reality is that most if not all hosting providers suffer from service interruptions. On the front end, the Web is this wonderfully, ethereally virtual place, but on the back end it’s filled with overheating boxes, coolant, electrical surges, and other deeply physical unpleasantness.

As brilliantly innovative as Web has made us, we are too often reduced to glitchy, overtaxed, or blacked out boxes on rack space in some remote location. Even Google, which has more infrastructure than god, makes the Web look like this sometimes:

Imagine how you would feel if you had a blackout in your home for an hour or two every day. But on the Web, we just accept that sites go down all the time and that there’s really nothing we can do.

As we increasingly satiate our unyielding appetite for video through IP instead of broadcast and cable, the problem is only going to get worse.

For now, I’ll be happy if I can find my own little oasis of uptime.