Imagine driving your big SUV around New York City looking for South and Broad, but due to a typo, you were supposed to find 220th and Broadway. You drove an extra 30 miles, wasted two hour’s time (without traffic), burned six gallons of gas, and emptied three tons of carbon monoxide into the air. What if this was a common occurrence, say three times a day, 365 days a year? 30,000 wasted miles. What if 100 people had this problem – 3,000,000 wasted miles? Because time is money, data is also money!
Suppose you run a courier company and instructions come via automated CRM software with the addresses for pickup and delivery – and that data is wrong (as above) 30% of the time, 3 orders out of 10, daily. That’s a real-world, time-wasting, energy-consuming, expensive problem. Let’s say you’re UPS or FedEx or a national bank, and you’ve got millions of queries into your database daily, and your database contains millions of duplicates and millions of incorrect entries. Huge problem!
Through efficient virtualization of s
ervers and optimization of configurations of networks, software and databases, companies that run large data centers are reducing their energy consumption. This means tens of millions of dollars in electric bills saved annually – for just one (1) large company. It means less demand on the aging, deficient energy grid. It also means hedging against future costs of power, because we don’t know what electricity will cost in five or 10 years. This means long-term sustainability.
But still it takes more CPU cycles and time for search and retrieval when the relevance of the data found is suspect. If your database is 30% more complex and 30% larger than it needs to be, because of incomplete, incorrect, inconsistent and duplicitous data, then it is 60% more energy consuming than it should be. That’s if you can limit your attempts to just one or two to find what you need.
The server virtualization benefit is decreased costs for servers, software licenses and networks, and the power consumption (CapEx and OpEx reductions) but – can this be improved further still? The answer is emphatically “YES”! Reduce and optimize your data – there is an exponential explosion of volumes of data, both structured and unstructured. But how?
Cleanse, consolidate, correct and complete the data, tag appropriately, then eliminate the duplicates. Also, employ intelligent data archiving / retention strategies. This reduces your database footprint. If the data is cut in half, the search time is cut by 4X (basic binomial search performance), thus the power consumption for each search (by human or machine) is cut four times. But there’s more.
Keeping the data “clean and green” via real-time data governance (and organizational data governance) and making sure newly added data conforms to data models and rule
s (business and environmental) helps assure that the power consumption levels stay reasonably linear or reducing, rather than increasing as business speeds along.
Bad data doesn’t just mean bad business, it also means bad for the environment. The new paradigm for data isn’t just clean, but green and lean.
An email has been sent to: