Microsoft Playing Nice in Open Source
May 21, 2012, Open Source Business Confrence, San Francisco—Sandy Gupta from Microsoft talked about “Mixed IT and Openness in the Cloud”. Interoperability across platforms and operating systems is becoming a necessity in today’s compute environment.
The cloud and open source are the intersection of business and technology. Cloud adoption is greatly exceeding all projections. New non-silo solutions are evolving because people are working together.
Some trends driving open source include: choice in solutions, agility, virtualization, and heterogeneous computing. Now there are many private cloud scenarios and interoperability solutions. The automation layer supports various operating systems and hardware platforms. Any provider has to work with companies and the various open source communities.
Enterprise IT now has to ensure that services meet the needs of that company. Virtualization and other cloud technologies are the key components. The overall system needs to have flexibility and freedom as well as quick responses to changes in the operational environment, driving interoperability up as a requirement. The systems need bridges between Windows and Linux.
Some vendors are now supplying management packages and desktops to automate tasks like provisioning. Integrated tools provide image of all facets of the system operations and health including the hypervisors and other systems functions. One vendor, Suse, is providing a management package for systems with extended operations management functions, cross platform views and reports, and views to patches and other maintenance issues.
The new environments force changes in relationships and partnerships. The public cloud is a playground for developers while the private clouds continue their journeymen work in keeping enterprises working. The volumes of data are growing exponentially and now users must be more aware of data types, storage, compute resources, and analytics.
One example of an open source, cloud-based tool is Hadoop. This software is designed to address the big data issues; size, velocity, and structure in an era where a search engine like Bing addresses 100 PB and gets 7.5 PB more every month. Hadoop uses a technique called map reduce and gives data a scientist focus for things like pattern mining, graph learning and other big data functions.
Big data is now the new oil, the data needs to be refined before it is useful. For example, 15 years of stock pricing data can take over 10 TB in Hadoop. It raises the question, what is your backup strategy for that much data. By creating a Hadoop cluster on Azure, they enabled an automatic refresh and update function and save to a Sharepoint account. The system can also generate reports. The workloads and cloud adoption require that all companies work together with the open source communities.
