This article aims to provide a more specific insight in to the most popular form of virtualization used within the industry: The Server Virtualization.

Server Virtualization, put simply, is a way of enabling multiple servers to be consolidated in to a single set of high performance hardware resources on a single physical server unit (or an aggregated hardware resource pool spanned across multiple physical servers known as a resource cluster) through the introduction of virtualization software. Virtualization software can sit on top of an existing Operating System (OS) or even directly on the hardware (bare metal virtualization) of the physical server (Host). It then presents carefully managed portions of the under laying hardware resources to the Virtual Servers (vServers) running above in the form of virtual hardware. This is carried out and managed by the virtualization software through complex and precise algorithm based simultaneous resource sharing making the maximum use out of the available physical hardware, while not degrading performance at the virtual server level.

In an enterprise server virtualization scenario, there will be a number of physical hosts attached to a common, shared storage (i.e. Storage Area Network) where all the vServer data is kept and is accessible by all hosts. This allows the vServers to run on any available host at any given time based on the resource availability. A simple Server virtualization solution would typically looks like below.

Server Virtualization Diagram

A virtual server can be created using the built-in tools provided within the virtualization software and then be built as per any other server. Alternatively, an existing physical server can be directly converted to a virtual server using the conversion tools provided by the virtualization software vendor. This is known as Physical-to-Virtual conversion (P2V). Once the vServer is up and running, these can be managed through the integrated tools available within the same virtualization software.

We will now talk about main benefits of server virtualization with some tangible evidence and then briefly mention the steps involved in carrying out a, sensible and money saving server virtualization exercise. We will then, in future articles look at how specific server virtualization technologies like VMware’s Virtual Infrastructure 3 and Microsoft’s Hyper-V based virtualization solutions helps you embrace server virtualization in order to reduce your IT operational costs.

In traditional IT, provisioning a physical server to host a server application has always been a time consuming and a trivial task. Selecting the make and the model of the server that is appropriate for the job and can be supported by the IT support team and also ensuring that a sufficient hardware maintenance agreement is in place with the limited budget you have is hard enough, but having to gauge the exact resource requirements (i.e. Memory and processing power) for the application and ensuring that there’s enough resource power for peek requirements based on the industry best practice could make it even more of a daunting task specially knowing that if the application underperforms due to incorrect resource estimates that it is your neck on the line. For this very reason it has been very common practise for most IT consultants, Server engineers and System administrators to almost always over provision resources to a server just for their piece of mind.

While this has been acceptable (or shall we say inevitable without a choice) for most of the non-IT savvy business managers, they have failed to realise a one key point. Over provisioning resources to servers just so that a sudden memory peak or a processor peak that happen once in a while can be catered for, during the other 95% of the time, that extra resources are not effectively being utilised. In today’s cost conscious business environments which are destined to drive their IT costs down, this is clearly something they could do without. Server virtualization eliminates this wastage through providing a single resource pool that can be shared amongst multiple servers where resources are dynamically allocated or de allocated as per their actual resource requirements at a given time.

Aside from above, another key benefit of server virtualization is the amount of cost savings it brings about through reduced energy consumption. As the number of physical server head count reduces through server virtualization, the energy you require to power up your data centre is also greatly reduced. Take for example, a scenario of virtualising a server farm of 45 ‘HP ProLiant DL380 G5’ servers and assume they all can be virtualized on a 15:1 ratio in to 3 physical servers (HP ProLiant DL580 G5) without any performance reduction. A DL380 G5 on average has 1000W maximum rated output wattage requirement and a DL580 G5 on average has 1200W maximum rated output wattage requirement. Assuming that each server is utilizing 100% of the maximum rated output wattage requirement at all times, the total energy savings from the above server virtualization scenario would be approximately as follows.

• Total Power requirements per server per year = 8,760KW

• Total power requirement per 45 servers per year = 394,200 KW

• At a flat rate of 0.14p per 1KW, total energy cost for all 45 servers per year = £ 55,188.00

• This is now replaced by 3, Physical servers costing = £4,415.04

• The TOTAL SAVINGS achieved through reduced power consumption ANNUALLY = £50,772.96

There will also be further cost reductions through reduced cooling requirements. Due to the reduced physical server headcount, the reduced heat dissipation within the server room now requires much less cooling power and this also ads on top of the above savings resulting in organizations achieving direct, tangible cost savings right from the start. Furthermore, Server Virtualization also brings the following major benefits to an organization.

• Significantly reduced hardware maintenance costs

• Dynamically expanding resource pools:-eliminate the need to over provision resources at every server level.

• Efficient utilization of their hardware resources: – get the best out of your buck.

• High availability and built in resiliency: – all vServers benefit from always on connectivity.

• Built-in Disaster Recovery (DR) capabilities

• Ability to provision new servers within minutes.

• Integrated management, monitoring and reporting of vServer health

• Built-in mechanism to support utility computing where server resources can be charged back based on the actual utilization:-Ideal for IT outsource companies or managed solution companies.

• Reduced Total Cost of Ownership (TCO) per server.

A sensible server virtualization exercise would have the following steps involved in setting it up. Please note that these are only high level, guideline steps not specific to any particular server virtualization technology.

1. Analysis of the current server farm: – Carried out through a ‘Capacity Planning’ exercise where the whole server farm is monitored using an appropriate capacity planning tool. This produces a report outlining what servers are ideal virtualization candidates and how many physical hosts that can be consolidated in to including any head room for resource peaks and vServer failovers. While this report provides details of exact resource utilization across all monitored servers, it also produces the approximate cost savings that can be achieved through energy usage reduction.

2. Evaluate multiple virtualization technologies and select the best virtualization software.

3. Ensure / procure sufficient hardware resources (Physical Hosts and Shared Storage) availability based on the capacity planning exercise and purchase the virtualization software

4. Build the infrastructure and perform relevant tests.

5. P2V the identified servers and shut down / decommission the old hardware / or re use as per appropriate

6. Document the solution and hand over to support.

Our website can be found at http://www.hcidigital.net and we always appreciate people following us on Twitter (http://twitter.com/hci_digital).

Thanks for reading

Chanaka Ekanayake
Network and Support Manager @ HCI Digital

Virtualization simply put is a way of separating your physical hardware from the system software i.e. the Operating System (OS) by introducing a new layer of virtualization software in between. This enables you to run multiple operating systems on a single piece of hardware. The virtualization software can either sit on top of the hardware directly, known Bare Metal Virtualization, or can even sit on top of an existing OS. It is seen as a major, if not the key, part of an overall trend in the industry that includes autonomic computing in which an IT environment is self managed and also in utility computing whereby computer resources i.e. processing power memory storage…etc are deemed as a utility which can be paid as per the actual usage. In its very early age virtualization was met with a great deal of resistance mainly from traditional IT pundits who either did not appreciate its worthy cause due to lack of knowledge about the new capabilities it brings to the table or due to their traditional scepticism to anything new and different. However things have changed a lot since then and some form of virtualization within IT is given its due place in today’s cost conscious, green label loving business management. To give you an example, every single company within the fortune 100 company list use at least one form of virtualization as of today.

There are different aspects of virtualization such as

• Server virtualization: The most widely adapted form of virtualization where by through the introduction of virtual servers as opposed to an expensive physical server organizations can reduce their physical server headcount dramatically while achieving more resiliency and fault tolerance within their server farm through built it technology within the solution itself.
• Desktop virtualization: Replacing a fully spec personal computer deployed at every user end with a virtual desktop that is running within a secure data centre with high availability (uptime). This reduces the computer cost per user by 1/3 or in some cases by half.
• Data centre virtualization: A fairly new addition to the virtualization arena but mainly used within data centres. Used commonly alongside server virtualization.
• Network virtualization: This is a way of implementing the available resources of a physical network by splitting up the available bandwidth in to channels each of which are independent from one another and can be assigned to different devices at the same time. Network virtualization masks the inherent complexities of a physical network from the user by separating it in to manageable components. Again used commonly with server and desktop virtualization.
• Storage virtualization: Most commonly utilized within SAN (Storage Area Networks) and is a form of presenting multiple storage devices as a single logical storage unit which can be managed centrally.

There are multiple technologies available to implement virtualization in the market. The most noticeable ones are from VMware, Microsoft and Citrix. All of their virtualization technologies address the subject differently but the concepts covered are all the same. While Citrix based virtualization has been around for a while, it is deemed not as flexible as its competitor’s solutions while Microsoft’s virtualization technologies, in our opinion, require a little more maturing. This puts VMware’s virtualization solutions right at the top in terms of popularity at the moment due to their maturity which explains why every one of the fortune 100 companies uses a VMware virtualization product within their organization.

Virtualization is a relatively large topic that cannot be covered within a single post but the attempt here has been to provide a very brief insight to the concept of virtualization as a whole and who offers what kind of information. There will be future articles published here which are more specific to different kinds of virtualizations with more in depth details and analysis of each vendor’s products so stay put…!

Our website can be found at http://www.hcidigital.net and we always appreciate people following us on Twitter (http://twitter.com/hci_digital).

Thanks for reading

Chanaka Ekanayake
Network and Support Manager @ HCI Digital

So, starting at the top, there are two things that people may talk about as being SEO and these are Natural SEO and Pay Per Click (PPC) advertising. PPC is the act of paying Google, Yahoo or most other search engines to put your small advert up when certain key words are searched for. These are not usually shown as part of the search results but as extras on the side or above. Some search engines (e.g. ask.com) partially merge the two but an experienced user is able to tell the difference immediately. PPC is useful and good to invest in but by no means does it out rank Natural SEO as the optimisation of choice.

Natural, sometimes called Organic, SEO is the act of managing your site and its presence on the web in order to be able to come near the top of search engines results page when something about your product, service or company is search for. It is commonly accepted that the first six results are the place to be and, if you believe some of the hype talked, nobody ever clicks page two or three. Of course they do really but not if something of interest has come up on page one.

So, how do you go about SEO and what do you look for in a company?

1) Search Terms

Knowing your search terms should be the first thing you look at or are asked about by a potential supplier. I was recently in a meeting with a company who were very happy that their site was number one for a particular search term and were very happy with their SEO supplier. However, and I don’t intend to criticise here just highlight, the search term was so specific that nobody was ever really likely to search for it in the first place and it was therefore an easy thing to do to get number one on the results. Nobody else was particularly trying to fill that spot!

The moral of this really is to look carefully at what terms your customers, not you, would look for in the search engine and be realistic about the response you get when you type it in. If you’re in the perfume industry for example, then you’ve got a challenge on your hands to get to the top of the list for “perfume” ahead of the other 65,000,000 results!

So, part one is to look at your search terms and your expectations.

2) Technical

This is the foundation of SEO. If your site isn’t put together in a way that allows the search engines to crawl your site and find the keywords you want to target then you are immediately at a disadvantage. We can’t go into too much detail here as we will fill a book but the basics are: JavaScript, Flash or Silverlight menus (bad, because the search engines cant crawl them), robots.txt (important, and need to allow your the search engines to crawl your pages, meta tags (important) and HTML tags (important, because they are how you highlight which words are important in the content H1 tags for the titles for example).

3) Content

As eluded to above, having your keywords (search terms) scattered around your site is a good thing. However, it will be better if they are nearer the top of the page and in bold, or preferably part of your title.

4) Management

This is really about managing your presence on the internet; you need to work to get links to your site from respected sites. A respected site is one that has been referenced (linked to) by a lot of other sites. You can do this through free advertising with the likes of Thompson Local or by paid advertising in yell, thetimes.com, yahoo directories etc…

There are a lot of good books where you can find further information on this or give us a call on 08454 67 67 18. Our website can be found at http://www.hcidigital.net and we always appreciate people following us on Twitter.

Look out for our new YouTube Channel (http://www.youtube.com/user/HCIDigital ) as we’ll be updating it shortly.

Thanks for reading and please feel free to comment…

But at any point was the user mentioned? The poor soul who is going to have to make the system do all the things the 400 page document says it will? I’m just playing around to some degree and to be fair there will normally be references in the document to the user. My point is that the business processes and requirements are usually used as the starting point and the user comes into it later, often driven by the graphics department / team. In your big document you’ll probably find a couple of statements of requirements 1) the system should be easily usable, i.e. have high usability and 2) the system should conform to a minimum accessibility standard of A.

There is another method of requirements gathering which is based on a process called User Centric Design (UCD), sometimes referred to as Interactive Design (ID). As its name suggests this process puts the user at the centre of the discovery / business analysis phase. This process, if done correctly, will bring about a more successful project as the whole process becomes geared towards understanding what the user needs to achieve on the system. Flowing from here comes the question “what business rules need to be followed in order to assist the user in achieving their goals?”

The difference here is quite subtle but important. The system will not fulfil its purpose successfully if its users are not satisfied with it. How many projects have been deemed a failure because its users have complained about the resulting software being difficult and slow to use? How many websites have you left because it was too slow or too confusing?

By putting the user at the centre of the project you will still have your business rules and requirements, you will still have these processes mapped and defined, but you will also have confidence that the systems designed are the systems that will empower people to carry out the tasks required. Whether they be a customer buying your produces, your finance department fulfilling the company shareholder responsibility or your knowledge workers accessing their required information.

For more information on UCD or any of our other services go to http://www.hcidigital.net or contact us directly on 08454 67 67 18.

You can also follow us on Twitter @ http://twitter.com/hci_digital

For medium to large projects, whether or not the development is in-house or through a supplier, it is an absolute must that somebody inside your organisation is overseeing the project and assessing the progress, milestones and risk from the business side. This can be an existing member of staff, a new employee or a consultant who works closely with you. They should have a good understanding of development projects as a whole and preferably the ability to question the development on a technical level; some people will still try to blind you with science.

So you’ve got your in-house manger / expert, do you now bring in the rest of the team? There are a number of considerations the first of which is that if you bring it all in-house, you take on all of the risk of the project. You cannot have a fixed price contract with yourself. You’ve also taken on the responsibility to ensure you specify the business requirements accurately so you will need to bring in a good business analyst who can work with you and help you build up a picture of what your business requires. There are the other usual risks but these are the two key ones that will affect your project.

A second consideration is whether or not you are going to want to carry out a number of projects over the coming months and years, if you are, then it’s quite possible that you would be better off building up a relationship with one or a number of key suppliers, so this next software project would be a good place to start.

The third key consideration is your deadline. Is there a business reason why you are aiming for that specific deadline that cannot be changed? Is the date event driven for example? If so you may not want to risk the learning curve that brining development in-house incurs and find out too late that you’re not going to hit that date. Unfortunately, some agencies may also tell you something is possible until you sign the contract then find reasons why it is not nearer the date. Unless you’ve got unlimited time then it is likely that you’ll have to compromise something, if it isn’t going to be the date then go for an agency with experienced leaders who will explain what can and can’t be done in the timescales. They would also be likely to look at phasing the delivery and talk through with you the scale of each phase.

There are other elements, some of which may be unique to your business, to choosing your approach than the ones above but these are the key factors you’ll need to consider. Most importantly, get somebody in-house who knows about the type of development your doing.

Hopefully this has been helpful but if you’ve any more questions take a look at http://www.hcidigital.net and get in touch. We’d also appreciate it if you’d join our followers network on http://twitter.com/hci_digital.

Do-it-yourself packages

Do-it-yourself packages are an option that need to be thought through carefully prior to putting your hand in your pocket. Choosing one of these “chose a template” options as your first point of call will generally lead to an application that is not user friendly, efficient or easily found. In these cases even if you manage to get some customers through this channel you will be unlikely to send them away feeling good about the purchase and telling all their friends.

If you are a small company or an individual and want to start in e-commerce without having to outlay thousands of pounds you will be getting an increased return on investment (ROI) should you initially build up your customer base through a channel like Ebay or Amazon. Once you have the custom, rethink your options from there, you could add a diy package as a small but growing market channel.

Bespoke build eCommerce

Developing an eCommerce application from scratch is the best way to get exactly what you want. You will be able to specify and decide your exact requirements and will neither be limited to the functionality of an existing application nor have to pay for functions that you and your company do not require. Having said that however, there are a number of risks to this approach and some rules that you would need to follow if the project would be successful.

It is, quite wrongly in our view, generally accepted that software projects will be delayed and over budget. This belief stems from the days when bespoke development was the only real way of going forward and they still carry an increased risk of this today.

The fundamental issue with this approach is what we call a “difference in world view” between the definer of the application (you) and the developer of the application (your supplier). Unless you define every potential user action and its response (which will lead to months of document writing and probably a delayed launch date) you will end up with a product that doesn’t do exactly what you expect it to. The supplier’s developers will often come across areas of functionality that are not clearly defined and due to the usual tight deadlines will make assumptions and complete the area to their specification rather than yours. Challenging the interpretation of a specified area at the end of a project is likely to delay your launch date, so staying closely involved in the week by week development of a project is very important.

A way around this is to agree to move away from the traditional waterfall project methodology and to use an agile approach, but this will bring risks of its own and won’t remove the necessity that you are closely involved in the project. We’ll talk about agile in another post soon.

So to sum up, if you have the time and experience to manage a bespoke build eCommerce development and know what you want, it is a good option and is likely to offer you an improved ROI.

Brand and Integrate existing products

There is a huge range of existing ecommerce products on the market that can be rebranded to be your new ecommerce site and they offer a wide range of functionality. It can be fairly straight forward to sit down with somebody who knows about eCommerce systems and list out the functionality that you need the system to supply.

Once you have your list you can almost always find an existing system that covers all the functionality that you need, though it may not always work exactly the way you expect, e.g. clicking add to basket adds the product and then takes you to the basket, whereas you wanted it to stay on the product page, or vice verca. This is where you have an option, take what you’re given or have it changed to work the way you want. Some changes will be straight forward and very quick, others could take weeks to implement and test. Again, the bespoke work is adding a level of risk but this approach limits your exposure to it, so long as you choose an experienced supplier.

Almost always there will be the need to integrate the application you buy into your existing systems and this can be one of the more challenging aspects of these projects, so if you can find one with a plug straight into your systems then great, but in reality expect some bespoke work here.

There are a number of other factors that will affect your decision of which system to invest in such as budget, expected usage and future plans. Implementing these systems can range from £30k to £3k so you have a lot to choose from.

Software as a Service (SaaS) – Renting eCommerce apps

Another approach to getting your e-commerce shop online is to rent an ecommerce package and plug your systems into it. This approach can make your costs more manageable as there is usually an upfront fee for integration and branding, then a monthly fee for use of the application. You will usually end up paying much the same as the other options over a period of 12 months so in the long run this can be a more expensive choice. The main drawback with these apps is that you usually get what you’re given and don’t have the choice of any bespoke work at all. But they can defiantly be a useful options and many leading brands have used them in the past.

We hope that this has been a useful read and if you need any more information take a look at our site http://www.hcidigital.net and get in touch.

In posts to come over the next few weeks we’ll be discussing the choice between in-house development and outsourcing, traditional project methodologies and the rise of agile and an overview of search marketing.

The current plan is that one of us will update this blog on at least a weekly basis with news, reviews and information so make sure you sign up to our RSS feed or check back frequently.

You can also find our micro-blog on Twitter at http://twitter.com/hci_digital