Marten Mickos takes on the cloud.

Marten Mickos, the former CEO of MySQL, has just been appointed the CEO of Eucalyptus Systems. I have a feeling that this means that many more people will become familiar with this company in a hurry.

Comments (1)

Back in Action: And Lookout Cloud

I has been so long since my last post I read it and it seemed like something that someone else had written. I quite enjoyed that. What really resonated for me was that just this morning I had a meeting about how we can work to have a cloud based product leverage the existing value that a CDN delivers. This meeting seemed like we were talking about something all new…that was until I read the article that I had posted on April, 09 of last year talking about just that.

This gave me mixed feelings on whether I was just ahead of myself last year, or were things just moving one year slower that I had expected. I think the latter is the most plausible.

All of that said I am making a commitment to keep more engaged in the cloud, and to be more active in the community.

Comments off

On Demand Cloud Migration Cost Analysis Tool

Again in my hunt for data about the cost of using cloud over being in-house, or using traditional hosting, I found what appears to be a very cool tool that will help you calculate the ROI of moving to the cloud. Now this application is specific to Amazon AWS it should provide solid ball parks for other vendors as well.

Check it out here: Aptio

Comments off

Interesting take on the Pros and Cons of Cloud

I found this concise posting about the pros and cons of cloud computing, while I have been working on a posting that will cover some of the same issues in detail.

Have a look: http://www.techsailor.com

Comments off

Combining Cloud Applications with CDNs Just Makes Sense

I my day to day work I am solely focused on providing solutions to my customers that will increase the performance of their web applications. Most people don’t need to see much data to understand that the better your application performs the better received/used/sales will be. Most organizations now have done about as much as they can, or know what needs to be done, to get top performance from their applications, and one item that is left to be addressed is latency. And this is where the CDNs can be the knights on white horses.

The combonation of CDN and Cloud just makes sense. You have all of the inherent value of moving to the cloud (scale, price, capacity at hand..) and the ability to move your content closer to your users. Well Mosso has really raised the bar in this space with their partnership with Limeight Networks with their Cloud Files offering. This is a huge leap in this space because it combines the leaders in two spaces; Mosso in Cloud, and Limelight in CDN. While other cloud offerings have been trying to simulate, or recreate CDNs, they have mostly failed due to the lack of proper infrastructure. Amazon for instance has one of the slowest time to first-byte times of any “CDN” out there.

This shows that Mosso understands that to offer solution to customers that in some cases you must partner with the best in the space. This also shows how, in my view, to use the cloud. There is not one vendor who makes ALL the needed tools, but there can be a vendor who ties them all together, and makes it easy to use every tool in the box!

This morning I read an interesting post on how to connect Cloud Files with the PHP API.

Comments off

Amazon picks up the game with Elastic MapReduce

This morning Amazon again took what some see as a leap out front with their announcement of Elastic MapReduce. They have utilized a hosted Hadoop framework across their Elastic Compute Cloud (EC2), and their Simple Storage Service (S3). This will allow developers to build applications that can process HUGE amounts of data, and do this cost effectively  by allowing Hardoop to automatically provisions, as well as reduce, capacity dynamically as load shifts.

This, trying to match capacity with load,  has been one of the main challenges with existing cloud offerings, and in my mind is really what makes the Cloud the Cloud.

I am really looking forward to looking further in to this new offering, gathering industry feedback, and presenting it back to you here.

Thanks,
Marc

Comments off

Open Cloud Manifesto Published

A concerted effort by several leading organizations, and spear-headed by IBM,  to outline a goal to keep standards, and keep them open, for cloud computing has been published. This documents main focus is on interoperability for cloud offerings. Over the past 60 years of computing we have seen costly lock-in when standards are not open, and defined. ”If the industry doesn’t come together, the result would be proprietary islands” of data and applications, warns Kristof Kloeckner, IBM’s chief technology officer for cloud computing.

The manifesto was circulated to leading organizations on March 22, 2009, and was quickly reviewed and agreed upon and endorsed by many, including AT&T, Cisco Systems, Sun Microsystems, Red Hat, and Advanced Micro Systems. That said however there are many heavy weights that have decided to step back.

Companies who chose to step back, at least for now, include Microsoft, Amazon, and Google (who pulled out after an initial endorsement). Steve Martin, of Microsoft, spoke to BBC about concerns over the timeframe, and the inability to provide feedback and suggest changes.

From what I can see out of my seat here is that the organizations that currently have the highest “cloud user-base” or are set to, are the ones who do not support the manifesto. Google, Amazon, and Microsoft seem to have the most to lose, or change.

One organization who also did not endorse the Manifesto is The Cloud Computing Interoperability Forum, who’s members include some who have endorse the Manifesto, as well as the key producer of the Manifesto (IBM).

This is a sure step that is going to force organizations to pick a side of the fence, for now.

Comments off

Fortune 100: Big Budget, no Cloud!

I spent the day the development, and architecture, team from a fortune 100 today. We had some great talks about everything from platform choice, to hardware refresh cycles. I found it very interesting that they consider moving some things to the cloud to have obvious value, but due to their security policies they will “never” be able to take advantage of such a move.

I was shocked to find that their SOP is a two year hardware refresh. And that goes for everything; servers, routers, laptops, monitors, firewalls…. and the most amazing part was that all of this is destroyed. Can you imagine how it must feel to crush say 25000 laptops, or the 15,000 servers that you brought on line just two years ago. They have projects that take over two years to roll-out or deploy, so in some cases even before something goes live they have refreshed the associated hardware.

Comments off

Why name this site ElasticStratus?

Elastic:  elastic rubber used to hold things (such as papers) together ; or holding applications/servers/companies

Stratus: low-level cloud type in the form of a gray layer with a rather uniform base, usually not associated with precipitation. Individual cloud elements have very ill-defined edges. Stratus is comes from the Latin word for “layer.”

This to me seems like a fit for what cloud computing is all about. A large layer with no defined edges, that can stretch to meet the needs of holding a growing item together.

Cloud computing has been something that has been of interest to me since I first was running into issues with scaling. When I think of how moving to the cloud can help with scale I am not just talking about how to scale your environment, but the many other components of business that also need to scale for success. These other areas are include your product offerings, versions of existing products, demographics for sales, launching entire new business with minimizing your risk capital.

When I talk to organizations, and peers, about the cloud there are very mixed reviews. Ranging from being completely against it, I think due to fear of perceived lack of control, being excited but not knowing how to move forward, or what would be wise to move to the cloud.

One issue, that I will dive deep into in a later post, is the issue of uptime and reliability. I think this is the least understood concept when talking about cloud computing. Well I should be more clear and say that the real issue is the inability for most people to accurately measure, and understand the difference between uptime and reliability, in both their datacenters and compare that to the cloud. I think this topic in-fact may warrant a full paper.

So welcome again, and I hope you come back often. You can also follow us on www.twitter.com @elasticstratus

Thanks,

Marc

Comments off