OPEN THOUGHT
DON SAMBANDARAKSA
According to IDC, Web 2.0-style collaboration is the No. 2 concern of CIOs today, second only to data centre reorganisation, and ahead of other hot items such as green IT, compliance or security.
Admittedly, that statement was taken somewhat out of context as it was a conference on green IT, but still, it is a wake-up call for many organisations who struggle to change from a paper and process based organisations with people working in the office pushing piles of paper from one desk to the next, into a knowledge organisation with staff scattered all over the world, working in their own time and space.
In the business of making news, images of Clark Kent and Lois Lane running around bumping into everyone in a mad frenzy to make the deadline are long gone.
Today, is is more likely to write-up in bed, at a McDonalds or even on a bus with a BlackBerry, removing the face-to-face contact and the social networks that come with it.
Clark Kent, Superman or not, would have known that Miss Lane was a great investigative reporter who would get her story come hell or high water, but would the two have fallen in love without the office scene?
In fact, of the strongest networks in an organisation is not one that can be drawn up officially, but it is the smokers, who have to find places to consume their legal narcotic together and talk to one another as their blood chemical levels stabilise. How do we then replicate this in cyberspace?
Email has been used to successfully maintain long distance relationships and networks. Maintain, that is, rather than build. It can work wonders for teams who know each other, but one cannot bump into someone on email.
Company email directories, the first attempt at fixing this problem, might say that there is a Mr O. Laden in human resources, but whether he is interested in flying planes into buildings or playing basketball after work, there is no way to know.
Yet, away from the office, social networks such as Facebook or Hi5 have made it so easy to find people with similar interests and connect with them. The good news is that big IT companies have taken note and are now providing Facebook style intranet solutions.
IBM has a product, Lotus Quickr, that addresses most of the pain points a typical 60-year-old organisation transitioning from paper to digital is facing. It has a blog section for the bosses to communicate with their minions, sorry, employees.
It has a welcome page for employees to see at a glance what every other department is up to. It has an innovation place where one can search for others with similar interest and pull them into workgroups. It has wikis for employees to contribute to employee-generated documentation. It also has the usual task management and document collaboration features and hooks into most legacy platforms.
SAP has a similar addition to its latest ERP system which looks extremely user-friendly and most un-SAP like when I saw it in Berlin this year. Rather than approach it from the user down, SAP seems to have built it on top of its robust, Germanic ERP and HR system. BT, the systems integrator formerly known as British Telecom, offers a similar "corporate facebook" built around Oracle-BEA's Aqualogic. The list goes on.
What the corporate-centred offerings have in common that is lacking in the free, non-work space is security and workflow. After all, a Facebook group discussing one's favourite politician is a place where everyone is equal. Whereas in a company, the boss does have the final say on matters simply because he is the boss.
"But we cannot afford to pay for this software," is a cry I hear all too often. Here again, IBM has a nice solution. One of the reasons it has developed the free office suite Lotus Symphony (circa 2008, not the circa 1984 version) is to free-up IT budgets from buying a commodity, in this case, Microsoft Office, and spend that money instead on IBM's consulting or, perhaps, groupware products.
Going a tad further, today we could get rid of even more of the Microsoft tax by running Linux in the office. No, I am not crazy, and yes, a lot of legacy Windows software is still out there that cannot be ripped out overnight. The solution would be to run Symphony (or OpenOffice for that matter) as well as other legacy Windows applications on virtual Windows machines in the data centre with the rich user interface running on any PC or thin client via Citrix's virtualisation software. This would save a lot of money, enough to buy Quickr and take the entire office out to lunch at a posher place than the office canteen. Or we could run a native Linux version of Symphony. Choice and freedom is so nice.
We would still run Windows, albeit on virtual machines, and in far fewer numbers than before.
The money saved is not just the Microsoft tax (Windows and Office), but also for taking care of and maintaining systems. Typically, the cost of maintaining systems far outstrips the purchase and energy costs of a PC.
Updating, patching and supporting a single virtual machine image that is actually in the data centre is much easier than running after hundreds of PCs on campus making sure they have up to date patches, anti-virus and the like. Compliance is also much easier with no real PCs out there to have a headache with. Plus, anyone using a MacBook Air can run the legacy Windows software too by opening a window to the virtual PC in the data centre.
The other common problems in transformation is a lack of a Project Management Office to coordinate projects (that is an episode in itself) and a tendency to repeat the mistakes of our forefathers and continue to build silos and stovepipes rather than leapfrog this unnecessary old-school thinking and go straight into cloud computing.
A stovepipe is a stack with storage, servers, application and, often, load balancing web servers on the top serving up the solution to the world. A modern CIO would know that it is wasteful and a nightmare when it comes to compliance and sustainability whereas an old school CIO would feel nice to see the storage and server physically there in front of him when he goes to work.
Instead of each application having its dedicated storage, everyone from EMC, Brocade to Hitachi Data Systems talks of having virtualised storage all pooled together. Legacy, older units can be used for slower tiers of storage whereas the new, faster drives can be added into a storage cloud and managed between different applications and severs or, ideally, applications running on virtual servers in the compute cloud. Sun likes to blur the line even more and offers blades with both storage and compute power in small, bite-sized easy to swallow units.
"But storage is cheap, why should I bother with all that confusing enterprise stuff you are talking about," is the response I usually get here.
No, hard drives are cheap, but storage is expensive, especially storage that can keep a CIO out of jail when it comes to compliance with data retention laws today.
In a typical siloed or stovepiped data centre, each item of information needs anywhere from five to 10 times its size to store it, or five to ten times waste. Take a typical content management system database project that is supposed to last for three years. On day zero, they will have to buy three years worth of storage immediately and create a database. The empty parts of the database have to be backed-up as well as the active parts. It has to be archived, wasting more space. Then the entire disk subsystem has to be mirrored to guard against hardware failure.
All of this extra work around the data should be part of the SAN (storage area network) and managed in the virtualised storage cloud. It should not be something that the application developer needs to concern himself with. However, unless there is a SAN cloud to begin with, and competent CIO, this cannot be done.
Ideally, the application should also run on virtual servers that run in the compute cloud. Web traffic from the US in the night can then use the same physical CPUs as the ones powering the virtual desktops that people write up their work on during the day, and storage can be added by the drive on a day to day basis at each day's declining prices rather than huge project outlays that depreciate faster than you can sing the national anthem.
Together, this maximises efficiency and makes everyone happy. Everyone, that is, except the old school CIO who fears losing control and the old school SI who is suddenly faced with a simpler, and less expensive project.
Prev
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Next
