If you looked into playing with Azure in the past, but did not jump in then it is time to take another look. Microsoft has added options over the last year that really remove objections to trying it out. If you have an MSDN subscription then you pretty much get a free playground in Azure that is going to waste if you don’t use it and if you don’t there is still the Introductory Special that goes through the end of March that gives you access to the basics of the service at no cost.
To look it over go to the Windows Azure Offers page at Microsoft.com and get going. You might not have a project that fits the Azure model currently, but you will. I am working on a new product for DTS that will have an Azure component and while it is still off in the horizon the time to jump in is before you are behind.
Over the years Microsoft has pitched alot of product and while I have always liked the technology (MS Bob, et. al. aside of course) they have not always been the most marketing savvy company when it comes to media. Over the weekend I what promises to be the best leverage of new media by Microsoft ever. The people responsible for pushing the MicrosoftWeb Platform have posted a video in the style of Magnum PI called Cannon PI (PI stands for Platform Installer) on YouTube.
Check it out and if you have the same reaction as I did you will be looking forward to where they take this series. It was an added perk to see Scott Guthrie and Soma playing supporting roles. Maybe that is why my interest is so keen?
In my business we deal with companies that are by their very nature risk averse and hence I only play with the newest tech for our internal projects, the occasional customer emergency and in my free time. Even so I have watched Microsoft’s Azure pretty closely and while I am confident that eventually we will take cloud computer for granted as we do dynamic web technologies now, I am also pretty sure that we still don’t know exactly what and how the real impact will take shape. Without clear SLAs and Pricing I just can’t gauge how reasonable it will be for a customer of size X with application of type Y to opt for Azure or any other cloud computing platform. That belief also drives me to think that the Open Cloud Manifesto is at best irrelevent and at worst a major impediment to getting where we want to go. If we don’t know what the best end state will be because we have yet to really evolve the technology in the real world then how can a group of people (any group) really hope to lay out the rules of the road. There isn’t a road built yet after all.
It has been proposed that guidance is needed to ensure that solutions are “open”. I can only assume that this means that they want code deployed on vendor A’s platform can be moved to vendor B’s platform unchanged (the classic case of wanting to not gamble on vendor lock in). While that is not specifically stated, I just don’t see any other interpretation that makes sense.
Time will tell, but I suspect we are several years of market testing and evolution from a point where we can even begin to have this conversation intelligently.
To read more on this topic I will point you to blog posts by Chris Auld and Michelle Bustamante. I must say that I agree with them for the most part.
StrangeLoop has finally announced their AppScaler device!
Richard Campbell told me about his involvement in StrangeLoop a while ago and I have been dying to tell people about it, but until now it has been confidential.
Basically the AppScaler takes a web farms major headaches and lifts them into the loadbalancer and out of the way of your developers. It really is a cool strategy because it gives sites real performance gains over hosting Session State on a state server or in a database along with a whole host of other performance enhancing and bandwidth saving features.
Check out the recent article at NetWorkWorld.com about it.
Sharing a web server between development teams is always fun (not). We had a problem surface today (or resurface) where if a developer creates a web application on IIS that uses .Net 1.1 for example (not an uncommon occurance) and some other developer creates a web application on that same server but this second one uses .Net 2.0 (something becoming more common every day). Odds are that the developers and even sometimes the network engineer or web master will allow the defaults to lull them into the false sense that it was an easy and straightforward task.
The problem is that they both allowed the “Default Application Pool” to remain selected and now the second of these sites to load will crash IIS.
You can’t have two different versions of .Net loaded into the same process and Application Pool often (though not always) means the same process.
Scott Forsyth has an article about this very issue that will help describe the error that occurs when you have this problem (the “Server Application Unavailable” error).
If you haven’t seen this yet, then you will.
I was recently in a discussion with some friends about Web 2.0 and what that all meant along with recently finishing a search for a hosting company for our dedicated servers. The two conversations actually led me to do alot of research and to alot of conclusions about the “next big thing” and how to get in front of it.
What follows are some opinions, advice and ruminations about the convergence of the two events:
As far as the next big thing, I think Web 2.0 is part of it. Web 2.0 being the idea that the web is great, but the real gold in this next round is taking data from the web and other sources and combining it in interesting ways that results in extended value. Like taking mapping data which is all the rage and tying into it for realtors so that prospective buyers can see not only maps, but see schools and stores and crime statistics. The data isn’t original as much as combined in an original way that adds value. It takes a mature Internet (Web 1.0) for Web 2.0 offerings to be practical and that is the age in which we live. When asked what the next big thing was my answers were things like robotics (especially in the military and law enforcement), commercial space operations, anti-biotic measures and other pharmacology rather than Internet or even computer technologies.
But when you think of it, you could build systems that are geared toward the above growth industries which is the traditional approach (that is where the high dollar demand will be) or we can combine data with functionality that wasn’t possible (or feasible) 2 years ago (ala the buzz word Web 2.0).
My point is that the two are related. You can revolutionize the horse drawn buggy whip industry with a Web 2.0 approach to data integration, but you won’t get rich on it. Another hazard is to not play where you will get crushed by the capital intensive crowd when the world sees you are making noise (think niche). The real danger lies in the lesson of experience in that every time I hear about a business proposal in an area that is not my field of expertise it sounds so easy (“what an opportunity”), but when you dig into the details you see all the complexities and barriers to entry under the skin.
Based on my own analysis, I am best suited either looking for a security offering to the growth industries or providing a value added service that pivots on security or finding a security product / service to build that leverages far flung data and innovative delivery (SOA).
Relative to the web hosting side of this, you have to take your idea and get it built without going broke and get traffic and attention without hitting the same obstacle (going broke).
In my searches for a reliable hosting company I found one that I like and that has worked well for us since we adopted it (we are starting to move more servers based on our initial success). The company is SRAWeb.Net which is very good on the Dedicated Server side of things. If you built your own Web 2.0 solution you could host it at a company like SRAWeb (or any other for that matter) and provide your service in the pilot phase on the cheap. I really like the idea of keeping start up costs as low as possible.
Once you get the process going, it is a treadmill and again if you haven’t done it, it really does sound easy until you dig into the details. Just get the search engines to send you traffic, all it takes is some SEO (Search Engine Optimization). I have made SEO a bit of a hobby and find sites like WebHostingFacts.Net very direct in their advice since it isn’t selling services (which are almost always overrated), but instead just gives advice.
The bottom line is that having a great idea is just the first step, it takes alot to get the get rich quick scheme to actually work. Luck is a mandatory and completely unpredictable ingredient as well. If you are upset that you missed that last big gravy train during the Dot Com bubble then think long and hard about the points I bring up here. The rumble of the next big thing is upon us.
I have started to encounter more and more instances where companies want to get out of the business of hosting websites themselves and since the price of outsourced web hosting has dropped the use of shared and dedicated server hosting has accelerated. There are many security as well as non-security related factors that should go into the decision on which approach is best for your application and I wanted to summarize them here.
I realize that to many people this is not news, but I am finding all too often that for a large part of the population this is a new insight so I intend to occasionally provide basic info as well as any advanced data that I can provide on both security and the growing practice of hosting sites with 3rd parties.
A 3rd party hosting company can afford to maintain servers at a fraction of the cost that anyone in any other business can manage. When you sign up for a web hosting company to put your site on their server you are typically looking at a very low price (under $20 per month and sometimes under $5 per month) and in these cases the web hosting company is actually not using a dedicated server for your site. If they were then they would be out of business very soon. The fact is that in this situation you are signing up for shared hosting and that means that your website might me one of dozens or even hundreds of other web sites hosted on that same server. The advantages of this model of course is the price. You could never get a dedicated server for the same price (they typically run over $100 per month for the bare bones package and can run into the thousands per month depending on the bells and whistles you require). The disadvantages of shared hosting are more numerous in my opinion than the advantages. On a Shared Hosting plan you cannot install any software that isn’t already part of the package, you might be sharing the same IP address as many other sites and the server is distinguishing requests by address once they arrive at the server, and most importantly if someone on the same server as you compromises the server with their web application (in the case of dynamic code) then your site is going to be dragged down too.
This isn’t an attempt to completely scare you off of Shared Hosting solutions, but be warned about the disadvantages before you jump at the price. I use shared hosting of simple sites that I consider low security, for everything else I go Dedicated Server all the way. I see efforts to save money or time as the most common sources of bad judgement calls that undermine security.