Storage at ETH
“Storage” was the motto of the forum for IT-users which took place yesterday at Hönggerberg. Approximately 70 visitors – mostly from an ETH-IT background – wanted to find out more about this current topic.
As the organizer, our goal is to offer an attractive information blend: presentation of existing services as well as new ideas or concepts – whereas this is not always easy to attain!
As introduction to the topic, a study by McKinsey and a small exercise based on Neptun sales figures allowed us a look at the wider picture which showed that the commercial world faces very similar issues and problems associated with storage (Meeting the demand for data storage, The McKinsey Quarterly, JUNE 2008, James M. Kaplan, Rishi Roy, and Rajesh Srinivasaraghavan).
The report on existing storage services made it clear that while ETH is well on track, the old idea of a “one size fits all” data storage is no longer valid. Cost reduction is not keeping pace with ever growing storage requirements. New compression techniques, such as Data Deduplication and entirely new approaches such as Wuala, will help to improve the situation, but ultimately it is up to data producers/owners to act cost-effectively. Only they can decide whether or not their data is scientifically valuable and irretrievable, and therefore need to be stored on expensive, highly available storage systems.
This decision requires a much closer cooperation between IT service providers and users in science and research. Lack of information between IT service providers and storage end users was identified as a major problem in the afore-mentioned study. The fact that very few end users attended this event confirms that in this respect as well we reflect the current trend.
In this sense, I appeal to our “regulars” with a request: Help motivate more of our colleagues from research to actively participate in events like this!