Virtualization

Virtualization, in computing, refers to the act of creating a virtual (rather than actual) version of something, including but not limited to a virtual computer hardware platform, operating system (OS), storage device, or computer network resources.

Virtualization began in 1960s mainframe computers as a method of logically dividing the mainframes’ resources for different applications. Since then, the meaning of the term has broadened.[1] 

… it continued and now this concept and technology is widely used to set up system architecture within data centers. Virtualized (data) servers populate physical servers. A well known company that is specialized around these questions is VMware.

Maps of data center localizations

Although data centers are unevenly distributed, but it’s intriguing to observe the way they’re located spatially. It’s difficult to find world maps but here are some examples I found interesting, but it’s totally not exhaustive (lots of them are not documented). Also note that any queries on image search engine about data center geography leads to weather-related visualizations (which generally influence energy/water consumption for this infrastructure).

The largest US data-centers (by Nicolas Rapp, data by Dave Drazen at Geo-Tel)

DATA_MAP_FULL

Thingful, search engine for data

Note: Thingful is a “search engine” (beta version) for data and artifacts/sensors that produce data (the coming “Internet of Things” so to say, but also and mainly weather stations, aircrafts and Rastracks at this day). Already quite loaded, the content of the search engine and its map will with no doubt explode in the close future. It is a project by former creators of Pachube (in particular Usman Haque), which was an open webservice to “store, share and discover” data from realtime sensors, now sold and therefore private… It was sold to LogMeIn in 2011 (which is somehow a sad destiny for an open data project, but this is a different story) and became then Xively.

 

thingful

Toward OwnCloud Core Processing Library

The purpose of the OwnCloud Core Processing Library is to give the possibility to program “cloud functionalities” within a well known and simplified designer oriented programming language (and community): Processing.

 

Therefore, the OwnCloud Core Processing Library linked with our personal cloud merges the Open Collaboration Service (OCS) Share API with higher level functions in order to implement seamlessly “search&share files” applications written in the well known designers oriented Processing programming language. This will soon become available to everyone on the I&IC website. The workshops we are currently running / will run during the coming weeks are helping/will help us fine tune its functionalities.

 

The OwnCloud Core Processing Library allows the automation of the action of sharing files and the action of file tagging within an open source OwnCloud environment. Search&Sharing tasks can be threaded and/or interdependent, everything depending on the kind of results expected from one application to another. Thus, these actions can be driven by unmanned processes, decision-making (copy, delete, share one or several files) based on related metadata (i.e. metadata relation/link) or based on external data, dug from the Internet or networked/connected items/things.

Setting up our own (small size) personal cloud infrastructure. Part #2, components

While setting up our own small size data center and cloud infrastructure, we’ve tried to exemplify the key constitutive ingredients of this type of computing infrastructure, as of November 2014. But we’ve also tried to maintain them as much open as we could, for further questioning, developments and transformations.

The first key ingredients are software parts and we’ve described them in the previous post about the same topic.

Comments on: Setting up our own (small size) personal cloud infrastructure. Part #1, components

Following Patrick’s post about our different options for choosing a “Cloud” software and the one that we finally made by choosing ownCloud. Here are a few related comments that develop our point of view and technical choices.

ArkOs, Openstack and RiakCS all take the hand over an entire server/system/computer, offering a kind of embedded linux system within a human-friendly interface, the kind of mechanism one can find on ready-to-use NAS (Network Attached Storage) hardware.

Basically, it transforms any regular computer into a NAS device. One of the key points about the structure we are trying to setup is to be able to host anything we would like/need or may appear interesting to probe. That includes our own website(s), web services in order to feed projects with data and any kind of applications that may be useful to try and develop within the frame of this research.

We do need therefore to keep the research server as generic as possible by using a normal linux distribution, which we can then enhance by any set of additional services. While ArkOS, Openstack and RiakCS are of course interesting projects, at some point, it may become already too specific for our goals.

 

iic_server_05m

 

Owncloud appears to be a simple web site structure dedicated to file sharing. As mentioned in my previous post, Owncloud proposes a set of APIs that allow the access to Owncloud features while being able to develop our own applications. Thus, these applications can rely on Owncloud while being hosted on a heterogeneous set of devices, network connected.