A lire sur: http://www.techrepublic.com/article/the-internet-of-things-will-force-companies-to-rethink-data-center-operations/#ftag=RSS56d97e7
By Michael Kassner April 16, 2014, 9:51 AM PST
By Michael Kassner April 16, 2014, 9:51 AM PST
The Internet of Things is destined to create huge amounts of data. Will data centers be able to stay ahead of the onslaught? Gartner analysts suggest things will have to change.
In preparation for the Gartner IT Infrastructure Operations & Data Center Summit held May 21-22, 2014 in Sydney, Fabrizio Biscotti, research director at Gartner, and Joe Skorupa, vice president and distinguished analyst at Gartner, teased the press world with interesting predictions about the Internet of Things (IoT). In a release, Biscotti started by saying, "IoT deployments will generate large quantities of data that need to be processed and analyzed in real time."
To paraphrase Biscotti, IoT devices will transmit real-time information to centralized management applications, which then will, with the aid of big data technology, provide an accurate assessment of how the parent business is functioning -- ultimately allowing management to make better decisions.
One example of IoT tech in use
Harley-Davidson has embraced the IoT, allowing its manufacturing plants to complete a new motorcycle every 86 seconds. In its recently renovated York, PA manufacturing facility, IoT sensors and associated applications keep track of production. This WSJ article quotes John Dansby II, vice-president of global manufacturing, "It allows us to be more consistent. In the past, operators had a bit of leeway on paint jobs, and each could do the work in a slightly different way."
What Dansby was referring to is how granular production management gets when using IoT technology -- painting in this case. There are sensors feeding temperature, humidity, and other relevant information into a management system that adjusts the appropriate parameter when existing conditions deviate from predetermined set points. As Dansby points out, "It is supposed to be an exact science, not art."
The challenges IoT presents data centers
In order to obtain the benefits from IoT technology, companies must be prepared for the blitz of unstructured data. Gartner alluded to this challenge in its press release. Data centers will feel the brunt of this challenge, since it will be the location where IoT data resides, gets manipulated, and are made useable. Skorupa added, "Data center managers will need to deploy more forward-looking capacity management in these areas to be able to proactively meet the business priorities associated with IoT."
The following areas are what Skorupa referred to:
- Security: The BYOD challenge will look simple compared to securing the myriad IoT devices connected to data centers.
- Enterprise: The big data created by IoT devices will present availability requirements, and an increased risk to business processes. Data centers will also become a more desirable bad-guy target.
- Consumer privacy: Gartner notes privacy issues will surface. For example, smart power-metering equipment will provide better service, but what guarantees are in place to insure user information will remain private as required by government regulations?
- Data: The most obvious challenge is where and how to store the enormous amount of IoT-generated data.
- Storage management: Storage alone is not enough. To be useful, IoT data needs cataloging for easy retrieval.
- Server technologies: The terabytes of IoT data will require additional servers and software, which in turn means larger support systems to handle increased electrical and cooling demands.
- Data-center network: Because of the huge influx in IoT traffic, WAN-link bandwidth will need to be increased.
An interesting conundrum
Recent computing trends have data located in a centralized repository. Skorupa feels that attempting to locate all IoT data at a single location is not technically or economically feasible. Skorupa continues, "The recent focus on centralizing applications to reduce costs and increase security is incompatible with the IoT. Organizations will be forced to aggregate data in multiple mini data centers where initial processing can occur. Relevant data will then be forwarded to a central site for additional processing."
If Skorupa is correct, this will mean additional challenges; disparate locations will have to appear as one, yet allow individual site monitoring and control. If this sounds familiar, it is. Biscotti said there will be more movement towards virtualization and cloud services. He added, "This will reduce the complexity and boost on-demand capacity to deliver reliability and business continuity."
Another challenge is data backup. How does one back up terabytes of data? Biscotti and Skorupa concluded that organizations will selectively back up data deemed valuable -- creating yet another conundrum. Trying to decide what data to back up, finding it, and ultimately backing it up will require additional processing, storage, and network resources.
If the IoT unfolds as expected, current data-center management, according to Biscotti, will have to use a Data Center Infrastructure Management approach, aligning IT and operational technology in a manner that effectively handles IoT data.