About hot and cold air flows (in data centers)

hot-aisle_cold-aisle_01

hot-aisle_cold-aisle_conditions

Both images taken from the website Green Data Center Design and Management /  “Data Center Design Consideration: Cooling” (03.2015). Source: http://macc.umich.edu.

ASHRAE is a “global society advancing human well-being through sustainable technology for the built environment”.

 

A typical question that arise with data centers is the need to cool down the overheating servers they contain. The more they will compute, the more they’ll heat, consume energy, but also will therefore be in need to be cooled down, so to stay in operation (wide range of operation would be between 10°C – 30°C). While the optimal server room temperature seem to be around 20-21°C, ~27°C for recent and professional machines (Google recommends 26.7°C).

The exact temperature of function is subject to discussion and depends on the hardware.

Yet, in every data center comes the question of air conditioning and air flow. In this case, it always revolves around the upper drawing (variations around this organization): 1° cold air aisles, floors or areas need to be created or maintained, where the servers will take their refreshing fluid and 2° hot air aisles, ceilings or areas need to be managed where the heated air will need to be released and extracted.

Second drawing shows that humidity is important as well depending on heat.

 

As hot air, inflated and lighter, naturally moves up while cold air goes down, many interesting and possibly natural air streams could be imagined around this air configuration …

 

OpenCloud (Academic Research) Mesh

Note: When we had to pick an open source cloud computing platform at the start of our research, we dug for some time to pick the one that would better match with our planned activities. We chose ownCloud and explained our choice in a previous post, so as some identified limitations linked to it. Early this year came this announcement by ownCloud that it will initiate “Global Interconnected Private Clouds for Universities and Researchers” (with early participants such has the CERN, ETHZ, SWITCH, TU-Berlin, University of Florida, University of Vienna, etc.) So it looks like we’ve picked the right open platform! Especially also because they are announcing a mesh layer on top of different clouds to provide common access across globally interconnected organizations.

This comforts us in our initial choice and the need to bridge it with the design community, especially as this new “mesh layer” is added to ownCloud, which was something missing when we started this project (from ownCloud version 7.0, this scalability became available though). It now certainly allows what we were looking for: a network of small and personal data centers. Now the question comes back to design: if personal data centers are not big undisclosed or distant facilities anymore, how could they look like? For what type of uses? If the personal applications are not “file sharing only” oriented, what could they become? For what kind of scenarios?

 

OpenCloudMesh_4.0_150dpi

(The reasons why an I&IC’s) OwnCloud Core Processing Library

Beside the reflection produced by the overall Inhabiting & Interfacing the Cloud(s) project and the related necessity to provide “access to tools” to a larger community (largely described in the founding document of the project and in a former post about the setting up of this library), new paradigms may arise in the global organization of servers farms. These new paradigms may in return generate new ways to organize files on cloud servers (by a different control of the redundancy principle for example, or a different use of file’s duplication, etc.), allowing for new projects.

In order to answer the stakes of the I&IC design research and to prepare such output/proposals, we have developed the OwnCloud Core Processing Library that will allow to setup a software layer on top of the hardware layer.

 

To download and learn how to use the OwnCloud Core Processing Library, we’ve prepared a post in the Cook Books section of this site.

 

owncloud_logo    processing2-logo

Cookbook > How to set up Processing to use the OwnCloud Core Processing Library

We will describe how to use the OwnCloud Core Processing Library within the Processing framework, starting from a blank sketch. Library’s functions will be refined and new ones may be developped, some additional libraries will be added as well in order to propose high level functions deeper linked to the IICloud(s) project.

 

own_processing_logo

Reblog > Power, Pollution and the Internet

Via The New York Times (via Computed·By)

—–

power_cloud

SANTA CLARA, Calif. — Jeff Rothschild’s machines at Facebook had a problem he knew he had to solve immediately. They were about to melt.

 

The company had been packing a 40-by-60-foot rental space here with racks of computer servers that were needed to store and process information from members’ accounts. The electricity pouring into the computers was overheating Ethernet sockets and other crucial components.

Thinking fast, Mr. Rothschild, the company’s engineering chief, took some employees on an expedition to buy every fan they could find — “We cleaned out all of the Walgreens in the area,” he said — to blast cool air at the equipment and prevent the Web site from going down.

That was in early 2006, when Facebook had a quaint 10 million or so users and the one main server site. Today, the information generated by nearly one billion people requires outsize versions of these facilities, called data centers, with rows and rows of servers spread over hundreds of thousands of square feet, and all with industrial cooling systems.

Mejias, U. A. (2013). Off the Network, The University of Minnesota Press.

IMG_9024

 

Note: Off the Network, a book by Ulises Ali Mejias that is interesting to read when it comes to objectify and question the network paradigm. Beyond the praise about participation and inclusiveness that was widely used by network advocates and now also by marketing companies, Off the Network brings a critical voice and addresses the centralization, or in some other cases the “nodocentrism” that is at work through many global online services, so as the commodification of many aspects of our lives that comes with them.

While we are looking for alternative “architectures” for cloud infrastructure, nodes and services, this is a “dissonant” point of view to take into account and a book that we are integrating into the I&IC bibliography.