When you hear “cloud computing” do you think it refers to something etherial or nebulous, appearing and disappearing as conditions change and difficult to get an understanding about?
Clouds are often a metaphor for the Internet “Moving to the Cloud” lets you use resources through the Internet that you don’t have available to you locally. (Wikipedia has a good definition and summary). So moving libraries to the cloud means moving library data and applications to the network level. Roy Tennant led off an overflow session with an overview of cloud computing. This has benefits: low barriers to entry, no capital investment because you can pay as you go, no local server capacity needed. Software upgrades are automatic, and you don’t need a dedicated staff. Of course everything has its drawbacks: you are giving up some control, and are relying on the network being available when you want to use it. Amazon Web Services has become a leader in cloud computing.
Libraries have been using cloud computing without realizing it when they used OCLC to create bibliographic records and then add value to them. The benefits of the information added by local catalogers thus become available to all. OCLC is a leader in cloud computing for libraries–click here for further information. Events and workshops on cloud computing for librarians are starting to appear, such as the WorldCat Hackathon, Mashed Libraries in the UK, and OCLC Code4Lib Bootcamp. These are working conferences; developers bring their laptops, collaborate, and produce working code. You can find WorldCat citations in Facebook, OCLC’s terminology services to enhance searches, visualizations of local holdings on a Google map, and WorldCat widgets for the WordPress blogging platform. All these are examples of the power off cloud computing when library data are moved to the Internet.
Andrew Pace said that moving to the cloud implies large-scale applications–you want many people to use them. The old model of users coming to libraries is not scalable, but then things like resource sharing, shared cataloging and consortial activities like journal licensing came along. These were the first attempts at attempting to scale library services. How can these be expanded? Repositories, shared discovery layers for catalogs, electronic journal management and link resolution are an entrance into the scalable library management in the cloud.
Libraries conduct about 166 billion transactions a year (about 5,000 per second). If you try to power this amount of transactions locally, you will soon need to buy more and more generators to supply power. When applications are moved to the network level, all you need to do is to plug into the wall to connect to the network, where the generators are backed up, maintained, cooled, centrally. Libraries have added more and more systems and made significant investments in infrastructure, but they still have a fragmented presence on the web. A webscale strategy will provide hardware and software on the Web, where they can use the applications they need. The potential benefits of this approach include user satisfaction and library visibility, workflow improvements, and financial savings.
Far from being nebulous, cloud computing is proving its worth and has generated some definite advantages. It’s worth looking into!
Columnist, Information Today and CIL2009 Blog Coordinator