It's the Year of the Dog

恭祝大家狗年旺旺旺!!

In translation, it means "wishing everyone in the year of the dog prosper prosper prosper"!!!

As a business venture, we will like to see the company prosper and grow. Though we must admit that the Information, Communication and Technology sector in Singapore is very competitive.

The number of entreprenuers has been on an uptrend in the recent years, fuelled partly due to more high tech startups and government initiatives, and partly due to employment squeeze here. Sometimes, we can be "reluctant" entrepreneurs.

According to a survey, managers and professionals in their mid-30s and 40s are most vulnerable in an economic downturn and are most likely to be retrenched. Once retrenched, difficult to find jobs of similar nature and has to settle for a "changing" career. Some ended up driving taxi...

Frankly, I do not consider becoming an entreprenuer when younger. I do not come from a well-do family, and would gladly settled for a stable, salaried post in some government-linked companies or multi-national corporations. However, 10 years later, I am not so sure anymore.

Standard ICT Operating Environment

The SOE (Standard Operating Environment) tender is coming... An estimated contract value of US$915 million is in the pipeline.

The Infomm Development Authority of Singapore (IDA) announced last year of its plan to create a standard desktop environment and a network environment across all government bodies, with the exception of Ministry of Defence.

Under the SOE model, external contractors own and operate the ICT environment for government organisations. The government will pay for the use of the services based on usage , similar to an utility model, on a monthly basis.

It is the latest initiative by the Singapore Government towards greater centralisation in the procurement of commodity IT products and services. The authorities believe that the SOE approach can help them save as much as 30% in annual ICT costs, by:

• reducing the complexitity of IT infrastructure and simplify its management
• increasing the speed and flexibility at which hardware, software and services can be rolled out
• lowering the total cost of ownership through economies of scale, eliminating hidden costs

This will effectively bring an end for Singapore Government bulk tenders of PCs and other IT products and services. But I am not so sure if such tender can create an even playing field to smaller service providers and suppliers, which may be marginalised in the view of economic of scale.

The winner bidders will have the contract for eight years, instead of existing three years government contract that comes with an option to renew it for another two years. So if a single vendor, likely a Government-Linked Enterprises or Multi-National Corporations, is awarded the tender, in the worst case scenario, will it dictate or even monopolize the SOE platform? This may results in less opportunities for the larger IT industry, especially the smaller players.

But one thing for sure, it look like Microsoft have a field day.

Tech Outlook 2006

The growth of IT spending in the Asia-Pacific region is expected to outstrip worldwide figures next year, while China will have the largest English-speaking population by 2008, predicts research firm Gartner.

Gartner projected that total global enterprise IT spending will reach US$1,768 billion in 2006, growing at compound annual growth rate (CAGR) of 4.5 percent from 2004 to 2009. In comparison, enterprise IT spending, which includes hardware, software, telecommunications and IT services, in the Asia-Pacific region will reach almost US$210 billion in 2006 at a CAGR of 7.5 percent over the same period.

Hopefully, our company can stand to gain from the bright technology outlook for 2006. The competition is cut-throat, and margins are getting thinner with the faster commoditization of technology products. Product life cycle is now measured in months.

It is a basic fact of life, for almost all industries in a free market, all products will be eventually reduced to a commodity over time. For information technology, it simply happens much more quickly. And it is never easy to patent technology, which is a common way to protect revenue stream and maximize profits.

A patent is an exclusive right to exploit (make, use, sell or import) an invention over a limited period of time (20 years from filling). The patenting of intellectual property involves substantial amount of costs, time and effort (averaging 2-3 years). In additional, patents are terroritorial, that is, effective only within the country where the application is made.

Patenting of technology is a double-edged sword, with different schools of thought on its effect on innovation, competition and technology diffusion. Open standards, on the other hand, allows for the wide adoption of new technologies in the marketplace. The potential for conflict between patents and standards arises when the implementation of the standard necessitates the use of technology protected by one or more patents. This has been one of reasons that is impeding the widespread adoption of open-source software such as Linux.

However, things are looking up for the coming year 2006.

Open Network Invention (OIN), a company founded in Nov 10, 2005 by IBM, Novell, Philips, Red Hat, and Sony, that plans to acquire patents and offer them royalty free to any company, institution or individual that agrees not to assert its patents against the Linux operating system or certain Linux-related applications.

Patents owned by Open Invention Network will then be available on a royalty-free basis to any company, institution or individual that agrees not to assert its patents against the Linux operating system or certain Linux-related applications.

Open Invention Network believes that creating a new system to manage and ensure access to key patents for the Linux operating system will have a significant economic impact. According to International Data Corporation, the worldwide Linux business is expected to grow 25.9 percent annually, doubling from $20 billion in 2005 to more than $40 billion in 2008.

Open Source Development Labs (OSDL), a non-profit corporation, founded in 2000 and has investment backing from Computer Associates, Fujitsu, Hitachi, Ltd., HP, IBM, Intel, NEC and others. The mission is "to be the recognized center-of-gravity for the Linux industry; a central body dedicated to accelerating the use of Linux for enterprise computing."

The Patent Commons Project was launched on Nov 15, 2005 by OSDL, an online patent common reference library aggregating and documenting information about patent-related pledges and other legal solutions directed at the open-source software community.

Monitor and Alert Systems III - Event Management

What is Event Management? Event Management can be defined as a set of functions that are applied to the alert system to identify those alerts associated with actual or potential service disruptions.

Event Manager (cont')

Alerts can arrive in different forms, as determined by the collectors and specific product implementations:-

• Simple Network Management Protocol (SNMP) traps
• Alerts from passive and active collectors using vendor-specific products
• Alerts from other elements using vendor-specific products
• Alerts triggered by the arrival and transformation of an extensible markup language (XML) document
• Alerts generated by the management system itself
  

Event management can be handled either by a real-time event handler or a policy-based event handler.

• A real-time event handler can determine problem priorities by following pre-programmed rules, and it can automatically activate other management tools. This enables the administrator to design a sequence of steps using each tool in the appropriate steps.
• A policy-based event handler is more sophisticated, as it applies business rules to the operation of the system. The policy manager is an automated tool that identifies the service levels allocated to each end user and application, based on a set of predefined rules.

Long Term Operations

Sometimes, operations are considered on a long term, because their activation or completion within a short time interval is not critical, but can be associated with strategic changes to the service-delivery environment.

• Stress Test - Through a set real tests of the system, or of a representative environment in a test bed, a load test helps determine the actual non-linearity points and bottlenecks within a system. This benchmarking information validates capacity planning and helps set the appropriate threshold for detecting problems.
• Performance Tuning and Capacity Planning - Using information collected over a period of time, these predict infrastructure usage trends and the resulting resource needs. Such data can provide enough time to get resources in place before there is any service-quality impact.
• Deployment of New Service - To ensure that the new services will not destabilize the current systems and cause service degradation for existing service mix, we need to prudent with applications rolling out into production. A lot of commercial software in the market are designed for functional testing, but more are also encompassing regression testing. When changes are made, a regression test ensures that the changes made in the current system do not affect the functionality of the existing product.

At this moment of time, we shall not go in-depth into each of these items. Hopefully, through the understanding the basic concept of "Monitoring and Alerts", you can put them to effective use with your solutions offering to your clients, using various complementary, cost-effective products.

Foreign Talents?

We are quite disappointed with the quality of the computer engineers from India. While I am not sure of what the rest of the folks in the industry think, must say most have pretty impressive resumes, deliberately crafted, sometimes deceptive. However, we observed that most seemingly lack the industry knowledge and/or hands-on experience to handle the practical aspects of things.

Nowadays, India look like a "factory" for mass production, with "assembly lines" chunking out large numbers of IT engineers to export overseas. Yes, they are cheaper to hire than our local chaps but when it comes to quality, we are often surprised by what we get.

Once, we are trying to install a software product on a Linux platform. We asked the developer, "What port does your program listen on?"

"Listen to port?", he replied, giving us a blank stare.

"If your program does not listen to a port, it is part of inetd?" we asked, because we want to more on the application flow.

"Inetd?"

We went on explaining to him how Internet services running on a Linux system, down to the concept of ports and sockets . This guy just cannot get it. We have to perform a "dissection" on his software to understand how his program works. Worst, he gave a lame excuse that the previous owner has not passed down the knowledge ... His manager of the same origns, is similarly clueless.

On another occasion, we are implementing Siebel CRM products at our customer site, the vendor sent down an impressive entourage of six "consultants". While I was on the system console, ready to install the software. Those six guys started poring over installation manual , crowding around me and looking over my shoulder.

"Ok, how do we proceed?" I asked.

Then they began, reading aloud from the manual the installation steps, discussing rowdily among themselves and suggesting what I should do next after every step. I got so fed up that I screwed them on the spot.

Why should we be engaging consultants, who come on site, stand behind you, read from the installation manual and ask you for advice instead of offering their professional services? We might as well read the installation guide ourselves.

To be fair, we expect foreign imports to the country to bring with them valuable experience and in-depth expertise. Because as a small business outfit, time is particularly precious, we cannot afford to become the training ground for these people. What appalling, it is not just the skills that is lacking, it is the fundamentals!!

No, we are not racists nor are we against our "foreign talent" policy. In fact, one of our founder is a local Indian. And we have worked with IT professionals from various nationalities and diverse background.

Perhaps, our friends from India has yet to realise the importance of good work ethics in business - a few rotten apples spoiling a whole barrel.

FreeRADIUS - Securing Wireless Connection

I should have ended with the FreeRADIUS trilogy. Anyway, we are implementing FreeRADIUS with wireless access point for a small business setup. Depending on the size of office, we with then cater for the number of wireless access points. But before rolling out the product at our customer site, we will first test out our solution. One wireless access point will be sufficient for the purpose.

(courtesy of www.linuxjournal.com)

A centralized FreeRADIUS will deployed allowing for authentication and encryption using the most up-to-date 802.11i/WAP2 Security Standard, which will provide the security level equivalent to that of a wired network.

More details can be found here:
http://www.linuxjournal.com/node/8017/print

Monitor and Alert Systems II - Instrumentation Management

In the last blog, we touched on the basic concept of instrumentation. For effective performance monitoring and fault management, system administrators will require related information on infrastructure behavior and the correlation between services and infrastructure usage.

Instrumentation Manager (cont')

The Instrumentation Manager helps identify the elements associated with any given service when there is a disruption.

Instrumentation Manager produces two primary outputs:

• Service Level Data that consists of data sets and aggregated measurements that are forwarded to the SLA statistics system for statistical treatment and reporting on system performance
• Alerts that resulted from data escalated to the real-time event handler, where they are combined with other data for evaluation

Thresholds and Alerts

Instrumentation Managers is used to configure the instrumentation systems, from which it receive the measurement of data. They examine each incoming data item, filtering out the obvious measurement errors and comparing measurements to specified thresholds to see if an alert should be issued.

Event Manager

If measurements indicate a possible problem, the instrumentation manager may demand additional measurements to help make sense of the problem and to see if the original measurement was an outlier or was a true indicator of a difficulty.

That will be the role of the Event Manager, which we will discuss in next blog:)