Spedi Retail Software - EPOS for the convenience sector

The World Wide Web Has Just Celebrated Its 20th Birthday!

Sunday 19th April 2009

In those 20 years, it has developed to an extent and in ways which could never have been imagined.

CERN, the European Organisation for Nuclear Research, is currently well-known for its high-profile "Large Hadron Collider" project and the associated problems. At the opposite end of the publicity spectrum, a very low-profile event occurred at CERN in 1989, which changed forever the way in which humans communicate and interact. Tim Berners-Lee handed a document to his supervisor Mike Sendall entitled "Information Management : a Proposal". Far from being a multi-billion dollar, groundbreaking project like the Collider, this was merely intended to suggest a more effective document management system for internal use by CERN personnel.

Then, as now, CERN was a large organisation, employing several thousand people, all of whom needed to share information in various forms. With the benefit of hindsight, it is not at all difficult to see how the Web could meet and even surpass this need. After all, it is by far one of the easiest-to-use and most comprehensive sources of information that the world has ever known. Miniaturised versions, sometimes referred to as "Intranets", are used by companies and organisations of all sizes to exchange information between staff, often on a global basis.

The Web's success, however, ultimately rests on Berners-Lee having realised that the key to the success of his proposal was the human interface side of browsing through this complex information space. He was mainly responsible for the early development of HTML, the mark-up language in which web pages are largely written. The actual browser running on the user's computer is irrelevant, as is the operating system which supports it. What matters is its ability to render the HTML and associated information in the required form. As a result, the Web is equally accessible to users running Firefox, Internet Explorer, Konqueror, Opera, Safari or indeed any other browser. In the same way, the operating system and server application running on the equipment from which the pages are being downloaded is totally unknown to the end-user. For various reasons, this is usually one flavour or another of Linux, with Apache as the actual serving application.

One of the main lessons to be learned from the success of the Web is that any proposed solution should be well thought-out, with proper consideration being given to the human-facing aspects. This need for "user-friendliness"can sometimes be overshadowed by the technical wizardry. When assessing "user-friendliness", even Windows is often seen as falling short of the Web, in spite of the huge efforts made by Microsoft's Human Interface departments.

Don't Neglect Your Backup Routines!

Sunday 12th April 2009

Otherwise, you will undoubtedly wish that you could rewrite history....

Taking a regular backup of any computer system is always a chore. In some ways, it is rather like the house insurance. Year after year after year, you pay the premium demanded by the insurance company and yet see no benefit in return. As yet another year passes without any need to make a claim, surely it is just a waste of money? Yes, it is - until the unwelcome day that a disaster strikes. It may come in the form of a particularly violent storm or major downpour, or even a burst in that massive trunk water main which, until now, you had not realised runs under the road right outside your house. No matter, the insurance policy should help to ease the financial pain.

Like the paying the annual insurance premium, backups often seem to be a total waste of time. It is important to remember that, just like the insurance premium, their true worth is apparent only when disaster strikes. There is a difference, however. Unlike insurance, the quality of a backup cannot be ascertained by reading the small print or asking questions. Arguably, there is only one situation worse than suffering a disaster when there are no backups and that situation is to have backups which are unusable. Simply because the backup routine copies the data to another medium does not mean that it can subsequently be restored, even in part. Equally, it does not mean that, even if they are able to be
restored completely, the backups actually contain everything necessary to replace what was lost. It is common practice to backup application data, but what about the applications themselves, never mind the operating system and all its labyrinthine configuration files? The question to be asked is "Were Server XYZ or Workstation ABC to fail completely beyond any form of recovery, could a replacement be obtained and would it be difficult (or even possible) to restore the backups to that replacement machine?"

Modern computer hardware is certainly very reliable indeed, but that does not help too much when a determined criminal has burgled the premises and stolen it.

The choice of medium used for backups varies depending on the environment concerned. Tapes, removable hard drives, DVD and Internet-based services are all readily available at varying costs. The point is that, once the backup method has been selected, it should be performed regularly. On an occasional basis, a test of the "What if...." scenario should be carried out. This will ensure that the backups are both sufficiently comprehensive and in working order.

Don't Cut Off Your Nose To Spite Your Face!

Sunday 5th April 2009

Times may be harder than they were, but that is all the more reason to invest wisely in IT.

Money is currently tight for many businesses. In such an environment, continued investment in computing hardware and software is often seen as an unnecessary luxury and thus an easy target to ease the financial pressure. Sometimes, this may indeed be true, but it is most unwise to slash the IT budget without a proper, in-depth consideration of the potential consequences.

When designed, specified and installed properly, new computing equipment, both hardware and software, can become an essential weapon in the battle for the company's survival. Depending on the circumstances, it may or may not provide a direct positive impact on the "bottom line". Instead, it may provide other, less tangible benefits, such as more timely and accurate business information, increased staff motivation due to less time being spent on repetitive, boring tasks and even open the way to entirely new lines of business.

One example of how investing in modern IT can help attract, motivate and retain staff is teleworking. Until recently, this was an option for only the most senior staff at the largest companies, due to the prohibitive cost of providing a T1/E1 or similar leased line, together with the equipment necessary to support it. Nowadays, reliable ADSL or cable broadband lines are available to over 80% of the UK population, at a fraction of the cost of a T1/E1 line. The equipment cost to support such lines is also very affordable. Equipment is, of course, necessary at the Head Office and security is a major consideration. Staff generally welcome the opportunity to telework, either partially or fully, especially when it is combined with any "flexitime" arrangements they may already enjoy. Very few people actually enjoy commuting, even if it is only for an hour or so door-to-door. The full or even partial elimination of the need for daily commuting instantly provides staff with an enhanced quality of life, to such an extent that it is almost invariably seen as being more attractive than an equivalent salary increase, quite apart from the saving on the cost of the season ticket.

It is important not to forget that times are hard for IT suppliers, just as for everybody else. As a result, a recession can be a good time to negotiate additional discounts, although not to such an extent that the supplier is himself driven out of business.

RAID - No Violence, Please, We're British!

Sunday 29th March 2009

What is RAID, anyway?

By any standards, modern hard disks are technological marvels. Their capacities are already massive and constantly increasing, although their physical size remains the same. Similarly, the speeds with which they store and retrieve the data is constantly rising. As for their reliability, many of the ranges of drives currently being manufactured by HitachiGST on behalf of IBM boast a "Mean Time Between Failure" of "Targeted 1.2 million hours", which is just over 135 years. Some of their ranges go even further and are described as being "Suitable for Mission-critical 24x7 applications". This, surely, must be "As good as it gets". One final feature of hard drives is that their cost-per-Gb is constantly reducing.

Unfortunately, there is always the risk of failure, as with any equipment. It is highly unlikely, although not entirely impossible, that even a brand-new drive may fail only a short time after installation. Although having a recent backup of the drive's contents is all well and good, there will still be considerable disruption to the users who rely on its services. This is where the various types of RAID can save the day.

The acronym "RAID" is generally understood to mean "Redundant Array of Inexpensive Disks" or, less commonly, "Redundant Array of Independent Disks". The two main types of "real" RAID are RAID-1 and RAID-5, with a number of sub-types. In addition, there is the so-called RAID-0.

RAID-1 employs the traditional "mirroring" approach. Two identical disks are used and all the data is written to both of them. Should one fail, it can quickly be replaced and the controller allowed to rebuilt the array "in the background", thus resulting in minimal downtime for the users. The downside is that only half of the total disk space is available, although there is one surprising benefit. Since, in an operational array, each disk is holding identical data, requests to read the data may be serviced by both disks simultaneously, thus increasing performance.

RAID-5, however, use a "striping" approach, where each block of data is spread across a minimum of three identical disks, along with a checksum. Should one disk fail, it may be replaced and the array rebuilt. This technique "wastes" only about one-third of the total capacity.

RAID-0 is not really a type of RAID at all, as it provides absolutely no protection for the data. In fact, precisely the reverse is true. Because each block of data is split between two disks, failure of one disk results in the unrecoverable loss of the entire contents of both disks. The MTBF, even with drives of IBM's unrivalled quality, drops to a little over 65 years, whilst with drives of lesser quality the spectre of total loss looms ominously over the data. The only benefit RAID-0 offers is that of increased speed, since data may be read from both disks simultaneously. Whether or not this increased speed justifies the risk is a matter for an individual's own judgement. After all, powerful RAID controllers and large fast hard drives with large on-board caches are no longer the expensive luxury they once were. If speed of access to data really is that important, why not simply use better drives in a RAID-1 or RAID-5 array, handled by a good-quality controller?

< Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Next >