Anyone know how I can find some cool things that happened in information technology on this day in history? I tried to Google it, but the results are useless. However, I did find out that today is Blog Action Day, which is an annual event held every October 15 that unites the world’s bloggers in posting about the same issue on the same day with the aim of sparking discussion around an issue of global importance.
From the site’s header, “Blog Action Day 2009 will be one of the largest-ever social change events on the web.” This year the topic is CLIMATE CHANGE. So here is my blog post about this topic, and I feel compelled to address this enormously important issue from a perspective of a load testing geek. How could I do otherwise? I am that man.
The way I see it, climate change is a result of people not taking load testing seriously. Nobody has ever load tested our planet to see if it’s infrastructure could handle the volume and maintain acceptable performance.
When the Earth was originally designed, did any of the stakeholders expect 6 billion system users?
Wikipedia states, “As of 15 October 2009, the Earth’s population is estimated by the United States Census Bureau to be 6.791 billion.” Climate change can be at least partially attributed to the impact of such a large load on the system. Each human consumes resources and produces waste. In the year 1 AD, the global population is estimated to have been 200 million, and in 1000 AD about 310 million. If the population had only grown by 100 million in the past 1,000 years, then I would expect a linear progression of resources consumed and waste produced. However that is not the case.
The acceleration of human population growth has been logarithmic, but the production of waste has exceeded population growth by an order of several magnitudes. It is logical that the impact on weather patterns would also be increasing at an accelerated pace.
It has only been in the past 100 years of so that serious data has been captured and studied regarding global climate. In some ways, that is too short a period (geologically speaking) to make a statistically sound conclusion about the long term direction of current weather data. Is the melting of ice near the North Pole a temporary blip on the Earth’s performance graph of response times and error rates? More load tests need to be run.
What metrics are predictive? No one exactly knows because no one built predictive models on the data we had 100 years ago. How do we select metrics that are useful in performance tuning the system? That is still under debate. Throughput, response times, concurrent users, error rate, requests per second are all good metrics for load testing, but are very difficult to apply to climate change. Most people use data like average temperature, snow melt on the poles, ocean level, and coral reef die-off rate.
Here are some stats I gleaned from the top Google result for the search “global warming facts”:
- Average temperatures have climbed 1.4 degrees Fahrenheit (0.8 degree Celsius) around the world since 1880
- The rate of warming is increasing. The 20th century’s last two decades were the hottest in 400 years and possibly the warmest for several millennia. And the United Nations’ Intergovernmental Panel on Climate Change (IPCC) reports that 11 of the past 12 years are among the dozen warmest since 1850.
- Arctic ice is rapidly disappearing, and the region may have its first completely ice-free summer by 2040 or earlier.
- Glaciers and mountain snows are rapidly melting — for example, Montana’s Glacier National Park now has only 27 glaciers, versus 150 in 1910.
- Coral reefs, which are highly sensitive to small changes in water temperature, suffered the worst bleaching — or die-off in response to stress — ever recorded in 1998, with some areas seeing bleach rates of 70 percent. Experts expect these sorts of events to increase in frequency and intensity in the next 50 years as sea temperatures rise.
- An upsurge in the amount of extreme weather events, such as wildfires, heat waves, and strong tropical storms, is also attributed in part to climate change by some experts.
The biggest problem with load testing the planetary environment is that we cannot replicate the production infrastructure; thus, any attempt at running stress tests and extrapolating the results to production will likely have a low probability of accuracy. This inability to accurate project the future of Earth’s climate make actual measurements from real-world test scenarios all that more important in our analysis. We must look deeply into our production system before we make an unsubstantiated jump in data modeling for the future.
Failures Unrelated to Load
???
Do we know empirically that climate change is attributable to heavy load? The data presented on the National Geographic Global Warming Fast Facts page is interesting, albeit inconclusive. I get uncomfortable with such statements as the one above about coral reefs. First, it isn’t a meaningful metric. It can be argued that there is no metric at all. Additionally there is no data presented that correlates the reef bleaching in 1998 to climate change. I am a certified diver and want to preserve every inch of every reef in the world! But I just can’t take the 1998 bleaching as a performance result attributable to load or volume relative to global warming. That seems to be an irresponsible use of statistics to me.
The last bullet point contains this phrase, “…attributed in part to climate change by some experts.” That isn’t helpful to a performance engineer. I would never make system tuning decisions based on someone’s third party reference to “in part” and “by some experts”. Those aren’t metrics that can be used for load testing, system tuning, or performance engineering. Those are more useful to journalists and Microsoft marketing managers that want to persuade you to buy Vista as a legitimate operating system.
Can we trace the system slow downs and performance goal failures to causes stemming from volume? I remember in the 4th grade at Deep Springs Elementary school hearing Mrs. Harding explain how the aerosol hairspray our mothers were using had caused a hole in the ozone layer. I have since noticed that pump bottles replaced aerosol cans, and freon in air conditioners cannot be recharged to get my car cold in the summer, and many other contributing chemicals have been found through scientific observation since then.
But I can’t help but think, “Is the climate change completely the fault of advanced civilization’s greedy ego-centric production of by-products and willful disregard for Mother Earth?” Could we simply be on a very long cycle of ice age thawing that is ramping up on the warming side? Put another way, are the performance results of our system related in some way to the normal degradation of processor chips and software application memory leaks? If one uses Microsoft as the leading world knowledge expert, we would be led to believe that this is normal and simply need to buy some upgrades to make things run better. Isn’t that a fair analogy to what United Nations’ scientists are telling us? If so, it would seem to me that it is time for the Bill Gate’s fix: re-boot then re-install the operating system then just buy another system.
“People will generally accept facts as truth only if the facts agree with what they already believe. ” – Andy Rooney