Cloud computing in laboratory and bio banking environments Kai Diercks, CEO soventec and software engineer by heart Cloud Computing what a word! For most people this term is really cloudy. It can incorporate all and nothing. A software developer would say: This technology is nothing really new. The technology is 15-20 years old, even most of the (web) services are that old. Is it really nothing new? And what is behind this big word, what is so interesting and why is cloud computing right now on the verge of a breakthrough? And especially what is in soventec interest: How is it useful in Laboratory Information Management Environments like < Lab OS >? Cloud Computing what is it? A definition for cloud computing is not that easy. The National Institute of Standards and Technology (NIST) tries to specify it as "Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction". To simplify it: Applications and services are no longer running on your local computer. You use services provided by remote computers and do not bother about the implementation or the capacity and capabilities of the computers providing these services. It is no longer necessary to have a powerful computer or even server. Technologies like CORBA, DCOM,.net Remoting, Webservices I won t go into detail hereor in general Remote Procedure Calls (RPCs) are at least 10 to 15 years old. The first ideas of RPCs have been published in 1976 by James E. White. Even web-based services like Microsoft Hotmail, GMX, Amazon or different Google web services are at least ten years old. These technologies are the basis of Cloud Computing! Why is now the time for the cloud and not already ten years earlier? In my personal eyes there are social and at least three technical reasons. There is a great acceptance and fascination for instant communication in private and business area with all its great possibilities in transparent and linked information and of course even the dangers. It is similar to the fascination when EMail or SMS has become commonly available in the beginning of the 1990ths. In private area a human basic need of social interchange is fulfilled although partially superficially - by these technologies, even if the person you talk to is in another city or on the other side of the world. It can be said much more about this, but it would blast this article. On the technical side the broad band networks have increased their capacities tremendously. They are low-priced and for everyone, everywhere available, even due to wireless technologies like 3G/UMTS. In the last 3-5 years the enormous developments on mobile phones have pushed trend. The mobile phone is no longer only a telephone. It is your universal communication tool, your navigator, your swiss knife for every life circumstance. Star Trek communicators have become real. Computer power is low-priced and Mr. Gordon
Moore ( Moore s law ) is right after all. When 1971 1 transistor cost about 5$, you will get about 1.000.000.000 transistors for the same price today. So datacenters like the Microsoft data center in Quincy, Washington US (13 soccer fields facility with over 30.000 real servers) or others by Google, Amazon or Yahoo are not a vision but reality today. These background factors push the cloud computing right now. In business field according to a representative study (XaaS-Check 2010) cloud computing is interesting for all companies, completely independent from the size of the company. Most of interest is the private cloud inside a company group (33%), only 5% deny cloud computing at all. But still 68% of the respondents worry about data security and availability of the services. My knowledge is only about the Microsoft data centers: Here the data or application is stored in minimum 3 redundant backups in electrical decoupled units. The data can even be synchronized to other continents in prevention of political or environmental problems. They guarantee an availability of 99,95% or you get your maintenance money back. I wonder, if any company server center can guarantee this Cloud Computing is it interesting for my project or company? Cloud Computing services are already available for many cases: Normal Office applications Data storage Data synchronizing etc. If you want to consume these services, it should be easy to find the right one. If you want to provide a service or build and companywide distributed application, the problem is more complex. But in each case you have to think about these points: Trust First of all you have to be aware, that your data is stored by another company. You have to trust them! But even if you mistrust them, analyze your data and decide, what is secret or could really be interesting for others. I would assume that this is only a small part of your data. You could store this part of data cryptographically secured or do not distribute this part of data. Costs Many consumable services are free of cost or you have to buy use licenses, you pay per use or if you cross a defined limit (storage capacity, band width ). The cost models are often very structured and easy. So a cost calculation should be easy. If you want to build a company application or provide a service on your own, you have to compare the costs, you need for the investment in installing and supporting the infrastructure for the application (administration, server hardware, supply costs), with the costs of using the cloud service. Often you have to plan your hardware according to peak usage, e.g. a mail server is definitely under stress, if all the employees come in the morning and checking there mails, but could be bored the whole day long. Cloud services can automatically be scaled up by using virtualization if more capacity is needed and scaled down, if less capacity is needed. Microsoft provides cost comparison online tools. In most cases you save between 30-80% of your costs. Ok, it is a tool
of Microsoft, you do not have to trust it at all, but it could be basis for your calculation. The Cloud will save costs if you have to instantiate a new company application, which needs extra infrastructure or if your IT infrastructure needs an update anyway. If you have a running state-of-the-art IT structure, you should have closer look, if working in the cloud in the right solution for you. Scenarios A typical application is the synchronization and transportation of data between different locations. For instance this could be a bio bank or laboratories which need to work on one data source but are distributed on different sites. Classical solutions (1) Data between the sites is synchronized via a secured VPN connection. The local data sources are synchronized via database sync services. - the infrastructure can be used for other applications /services Disadvantages - sometimes slow VPN connection - Server infrastructure has to be available on each side. Preventive measures for availability have to be provided (emergency power supply, backups, clustering) - eventually two data sources, synchronization problems can occur (2) The Data application is provided via a Web Frontend - one data source Disadvantages - web application often cannot offer complex functionalities like desktop applications - Server infrastructure has to be available and administrated. Preventive measures for
availability have to be provided (emergency power supply, backups, clustering) Cloud based solutions (1) Database servers are in the Data Center Remark: In this scenario another optional database is shown. It can be used for critical data, but can also be omitted. - one data source, redundant storage inside the data center - Backups, updates, availability, security is managed by the data center - usage is scaled according to needs - avoiding costs for server infrastructure Disadvantages - data is not at home
(2) Database servers and application is running in data center - one data source, redundant storage inside the data center - Backups, updates, availability, security is managed by the data center - usage is scaled according to needs - avoiding costs for server infrastructure - centralized license management for the application Disadvantages: - data is not at home - application is not at home, has to be administrated remotely Cloud Computing - < Lab OS > is ready for the cloud! soventec < Lab OS > uses future-proof technologies. Because of its internal architecture the < Lab OS > system can be integrated in cloud scenarios. It supports web services and other technologies to connect existing infrastructure, < Lab OS Sample DB> locally or as distributed databases in a datacenter or on different private sites, < Lab OS Mobile> (coming soon) connecting mobile devices like Windows Phone 7 or Laptops of mobile laboratories to the < Lab OS > System. The system integration and the components you need depend on your concrete project
or requirements. So feel free to ask! http://www.soventec.com