Friday, February 15, 2008

A Day of Storage and the Cloud

My reading began this morning with Nick's covereage of the Amazon S3/EC2/AWS outage. Perhaps most interesting to me, though, were the comments. A variety of people responded to note that we perhaps are holding the cloud to impossibly high standards, while others noted that this was supposed to be a distributed service, and an extended downtime like this indicates a certain lack of redundancy. I find this facinating, in light of the recent discussion of cloud lock-in. Not surprising, just facinating.

Let me explain.

While I remain extremely concerned about the proprietary operational approaches taken by most "capacity-on-demand" providers--many based on open source platforms, ironically--I think it is important to acknowledge that:
  1. 100% uptime is unreasonable for any platform in its infancy, including S3
  2. Not everyone will be negatively impacted as much by a three hour outage as some
  3. SLAs should be set if service is extremely critical to a business. Ironically, Amazon has limits on who and what they will provide SLAs for.
  4. Even with a three hour outage, Amazon S3 is probably the best service of its kind...for now.

That last point is critical, as Nick put up another post later in the day highlighting EMC's plans to enter the cloud storage market---in a big way. The competitors to Amazon are coming, and that fact may change the equation for how much leeway Amazon has in the future.

Assuming it is not super onerous to copy data from one provider to another--Storage may in fact be the earliest of the commodity cloud components if this is true--an alternative approach will make it that much simpler for an unsatisfied customer to make a move. This, in turn, will make some who will tolerate an outage now, well, less tolerant.

I anxiously await Amazon's explaination for the glitch.

By the way, Robert Scoble certainly believes Amazon has won the cloud market in its entirety already. He is way off, of course. Do you know how much datacenter capacity there is in corporate America alone? There is no way one company that is spending a fraction of the budget on building new data centers that Microsoft, Google and Yahoo are will create a barrier of entry that high. Amazon is a typical first enterant, ala Netscape. Hopefully the market is different enough, though, that they can build a survivor.

1 comment:

Anonymous said...

James,

This topic has had a lot of coverage since the turn of the year and Nick Carr's book gave the debate a boost.

In order to discuss some of the issues surrounding The Cloud concept, I think it is important to place it in historical context, looking at the Cloud's forerunners and the problems they encountered before being adopted.

On my blog, www.KeystonesAndRivets.com, I've tried to do that in my "Cloud Computing" post and I've linked to you and some others.

One of the current barriers in the way of The Cloud is economics. I argue that,

"Telecom prices have fallen and bandwidth has increased, but more slowly than processing power, leaving the economics worse than in 2003".

And, “I'm sure that advances will appear over the coming years to bring us closer, but at the moment there are too many issues and costs with network traffic and data movements to allow it to happen for all but select processor intensive applications, such as image rendering and finite modelling.”

Your comments and feedback are very welcome.

Regards

PJW