EpicStar SuperPost: Faith in the Cloud – the missing link

So here we are, LGEO Research’s first guest blog post! Hooray! Its such a pleasure for me to host this post by Andrew Beeken, a good friend of mine on Twitter who’s doing fabulous web whizzy stuff at Lincoln City Council. This post talks interestingly enough about a topic I’ll be debating at my upcoming event in Brum next week (details of event here).

Andrew highlights a very important challenge in embracing cloud computing and a great platform to start discussing its strengths, weaknesses and possible solutions. I have covered a talk by best-selling author @WeThink Charlie Leadbeater last year on how cloud computing will change culture and politics here.

You will find Andrew arguing some simple and easy yet revealing observations on how governments adoption of cloud computing to cut costs can impact services. Its a very interesting read and I highly recommend it. It certainly helps me to cover more techy stuff on this blog, as I’m a geek who’s techy shy 😉

I hope you enjoy it!

Liz xxx

By Andrew Beeken

I’ve called this post Faith in the cloud, as it shares similarities with a couple of other posts on my blog, StudioAB. You can find those posts here and here. Big thanks to Liz for hosting this post on her blog!

Anyway; here are the facts folks. At approximately 1630 hours today (Wednesday 15th September 2010), I noticed that I couldn’t access Google from my work PC. At all. No Google.com, no Youtube, no Gmail. Nadda. At first I just thought “Oh, something’s up in IT” until I realised that other websites seemed fine. So, in my puzzlement, I turned to old faithful. Response was surprisingly mixed – some folks could access it. Some couldn’t. Oh my. Google appeared to be having problems and was potentially down! Proper end of the world stuff! Cats and dogs living together; mass hysteria!

A while later and a bit more digging on Twitter turned up something interesting: the people who couldn’t access Google related services were all on a BT connection. Turns out there was an issue with BT’s DNS servers that was causing the problem. The issue was resolved in half an hour and everything went back to normal; the world didn’t end and we all kept on living. Hooray!

So, what is the issue here? Why do I feel compelled to write about the situation? Well, one of the things that interests me, and where some of my work lies, is in looking at the ways that cloud based tools can be used to deliver a better service across websites which are restricted by either budget or technology. Google’s tools feature heavily in that work.

Now, the immediate effect of the BT/Google downtime was comparatively minor; some people couldn’t access Google search, GMail was unavailable etc. That was all resolved, but when you think more about the problem it does in fact go deeper than the immediate Google services. Let’s take, for example, sites that rely on Google Analytics. Many BT users reported that the Analytics code trying to access Google was slowing those sites down, often to the point of being unusable. The same stands for sites using Google Custom Search, Google Maps and other Google related API’s. All sites that were not Google, but relied on Google’s tools.

BT recently announced that it had signed up fifteen million customers to it’s broadband network

That’s not only customers through the standard BT retail brand, but also through other providers using BT’s network – in other words a heck of a lot of people who, for thirty minutes of today, could not access a rather significant chunk of the internet. Thirty minutes doesn’t seem a massive amount of time, but for some online retailers, for example, it equates to lost sales. Of course, this is all very speculative – it’s almost impossible to quantify these hypothetical lost sales and point the finger of blame solely at an ISP, but the concern is definately there. What if this lasted longer than thirty minutes? What of the further reaching issues as companies like Google entice us with more rich services to improve both our online experience and that of our users and customers? Should we really be putting all our faith in tools like these when one simple link out of place can render the whole chain useless?

My answer is, as always, yes we should – but we should always have contingencies. We should embrace the luxuries and experiences that these services provide but we can never assume that they will be around or free forever; this has been shown recently with services like Ning and Hootsuite moving their freemium services to paid-for models. We should always have a back up plan and, most importantly, we should always have our data! Regular backup and export plans are essential to ensure that your information remains that way!

So what are your thoughts? Is cloud computing the future; something we should accept and embrace with all it’s potholes and foibles? Or should we be focusing on providing more solid, less connected functionality for our online services?