Hosting 365

That’s a moot point for a hosting company obviously - if the company HAS infrastructure of course it will boast about it, if not, don’t expect a detailed explanation of the Data Centre facilities on their web site.

But what about web applications and business or community web sites?

Would it foster more trust if you published information about the hosted services you have? Is that relevant at all?

I actually think it is - depending on the site. Basic sites that serve content to people, no matter how busy, will require good hosting, but to the people browsing the site that’s irrelevant. However communities, e-commerce sites, business and especially technology forums - these users do care about the web site and having some information about what infrastructure it’s hosted on would not only provide a better sense of security and reliability to existing users, but also help provide a measure of ‘trust’ for new users.

The same goes for web applications - it’s unlikely I am going to store any sensitive data (emails, credit card details, files / photos) or any data that I want to be able to gain access to, securely, at any time, from any where (that’s the basic tenet of a web app in fairness), with a web application provider that I do not have some level of trust in. Publishing some details (of course I’m not suggesting publishing anything that might actually give away too much info and have the opposite security effect!) of the hosting platform the application is served from, would give SOME peace of mind (of course financial and competitive feasibility are business considerations as well).

In all this rambling, to try and distill it, what I’m really saying is that any web site that has a user base (not a people-who-browse-only base) should publish data security, network connectivity, and server/platform resilience and redundancy information within their ‘about’ section. This is of course, like Hosting Companies, provided that the site in question is actually hosted in and on some good infrastructure!

If there’s any interest in this I’m happy to write and publish (under a Creative Commons license) some white-label data sheets and information on our data centre and network infrastructure that you (you, being hosted in here of course!) can edit, cut+paste, butcher and re-publish on parts of your own web site.

HP have made big announcements over the past few weeks about consolidating their many data centres and operational locations into a far smaller number. One has to wonder why HP, among others, are centralising all their infrastructure?

Clearly it makes financial sense to reduce overhead while maintaining or improving performance - which is the goal of these types of infrastructure consolidation.

The fact that HP, one of the largest IT companies in the world, is embracing this strategy, is an endorsement of one of our key messages - hosted applications.

Take a standard business with a couple of in-house servers as an example. The servers probably act as a file store and email platform. There is most likely a tape backup solution that the MD or resident IT-guru (who is full-time in a sales or admin role!) manually runs the backup, rotates the tapes and brings one home every evening; and there is a maintenance contract in place which guarantees repair / replacement SLA’s.

The consolidation strategy is one of reducing un-needed infrastructure. For a small business to have their own servers and most likely no full-time IT staff is lunacy. For starters the capital cost of hardware and software licenses is a lot, as well as the maintenance contracts and internal labour required for bug fixes and software patching.

We are trying to educate the market to the hosted application alternative. In this scenario, an Internet Infrastructure Provider supplies the machines, software licenses (hurray for SPLA!), power and IP connectivity. The IIP also guarantees better uptime and repair SLA’s than is possible in-house. Furthermore the backup can be automated and sync’d to an off-site (another data centre) location. The real carrot of this solution then is not only that it is demonstrably far superior, it is also massively cheaper - as there is no capital outlay, and only a minimal monthly spend (our Financial Controller - Ken Pierce - has a good Cost-Benefit comparision analysis on this … which we’ll hopefully make into a web-enabled tool soon).

Clearly the industry believes hosted applications (Software as a Service - SaaS - some call it) are the way forward - with Microsoft (whom long derided anything not based on a local PC) embracing it with their SPLA licensing scheme.

The greatest challenge is in educating the market - moving visibly equipment off-site has a perception of risk, even though it’s actually being moved to a better facility!

Gartner reports (no link, sorry, read it on REAL paper!) that the role of the CIO is to change from being predominantly internal in focus to a more external outlook, challenged with helping company competitiveness and growth.

The report peaked MY interest because it 2 of the technologies it mentions that will be key in 2006 are ‘mobile workforce enablement’ and ‘collaboration technology’. Guess what? They’re both hosted applications, or ones that hosting will play the part of facilitator and infrastructure provider to.

Collaboration can only really be achieved thru on-line applications. The days of the shared network drive are dying (thank goodness!) and hosted applications are providing a variety of functionality (enterprise management, customer relations, logistics - now even desktop applications like email, word processing and project management). Technologies such as blogs and wiki’s allow companies to communicate and share project / product / business information internally or externally. The primary benefit of these is access-anywhere, and edit without local application requirements.

Which leads nicely to the second technology - mobile workforce enablement. The CIO’s role will be to provide the hardware and software to facilitate an increasing mobile workforce. It’s no longer only sales reps and maintenance workers that are out-of-office, teleworking, offices at multiple locations (and outsourced offices or managed services providers located off-site) and just keeping in touch with the office out-of-hours. This means providing services like push-email, portable phone numbers, and access to what were traditionally in-office applications like file servers, CRM and Enterprise software.
On the business side, the report states that one of the key 2006 priorities will be controlling enterprise operating costs. Music to MY ears! The model we take allows CIO’s to procure a high-end cluster of dedicated servers with hosted office applications, disaster recovery options and managed services support (or any customisation of that of course, up or down scale), for a monthly fee that is a fraction of the captial expenditure normally required. That model saves CIO’s both cap-ex costs, and internal staff resources. Hey, of course we make money off it to, it’s the outsourcing business model - providing Internet Infrastrucutre and associated services are OUR core features, not yours, so why try to build it yourself when we’ve invested millions in people and facilities? Just like we don’t try to write our own operating systems for our servers, we’re quite happy to use best-in-breed systems that already exist!

I think hosting companies will move aggressively into this space - building on their data centre infrastructure to provide value-added services and pre-packaged solutions. It’s a win-win for both customers and the hosting company as they have the expertise and facilities to provide the hosted applications CIO’s will need to enable the key drivers of changes mentioned above.

There’s been lots of talk in Bubble 1.0 and Bubble 2.0 (that’s now!) about SaaS : Software as a Service (i.e salesforce.com, Microsoft Live), and it seems now that business, software and infrastructure have caught up with the IDEA of it, SaaS companies are becoming successful.

I came across an article by Nicolas Carr (of ‘Does IT Matter‘ fame) called ‘Here comes HaaS‘ and it strikes me that the dedicated servers model we use (and many other data centres) is exactly this.

Google Define doesn’t have a technical definition for it, so I’ll explain it here in plain English, or I’ll try! Rather than purchasing a server, installing your OS and other applications and connecting it to the Internet (all of which involve capital expenditure, and are on-going labour intensive), HaaS means for a monthly payment that is a fraction of your normal cost, a business can lease a server, with managed services support, standard OSes and bespoke application deployments, and a boat-load of connectivity.

For example, we sell a dedicated server - your own machine, which you can do with as you please (within reason on course, we have an acceptable use policy), install OSes and applications, and provides you with fast (and lots of) bandwidth - for only 59.95 a month! When I joined here first I couldn’t believe that, if I add the cost of buying my own computer and a simple broadband connection, if would be more than this a month, and my own machine in a data centre for this cost - WOW! Of course NOW saying that makes it look like I’m pimping our own products!
What’s important to know about Hardware as a Service is that it’s really relevant to you. Why? It takes away pain, it saves money, it provides access to dedicated resources (people and infrastructure). Moreover, I believe Haas is the enabler that will change the face of Software-as-a-Service. With Microsoft’s SPLA licensing, which effectively makes access to high-end software a lot more affordable, and of course, the Open Source OSes and applications, using the HaaS model, a relatively small monthly fee can provide you with office server funcationality, for literally a small fraction of the cost doing it in-house would be.

Watch the HaaS space, I think it will evolve into a H+SaaS model where bundled solutions will be offered rather than just empty-shell machines. There’s a business opportunity here for software companies to package and license there applications in the H+S-aaS model, and charge on a per-user / per-domain basis.